24 Hills Road, P 01223 855346 Cambridge W elifesciences.org CB2 1JP T @elife_sciences FOR PEER REVIEW - CONFIDENTIAL

Neuroanatomy of a hydrothermal vent shrimp provides insights into the evolution of crustacean integrative brain centers

Tracking no: 09-04-2019-RA-eLife-47550

Impact statement: Using a suite of state-of-the-art neuroanatomical techniques we provide insights into these animal’s brain architecture to illustrate possible adaptations to the hydrothermal vent habitat with its extreme physicochemical conditions.

Competing interests: No competing interests declared

Author contributions: Julia Machon: Conceptualization; Data curation; Formal analysis; Validation; Investigation; Visualization; Writing—original draft; Project administration; Writing—review and editing Jakob Krieger: Formal analysis; Investigation; Visualization; Methodology; Writing—original draft; Writing—review and editing Rebecca Meth: Data curation; Formal analysis; Writing—review and editing Magali Zbinden: Conceptualization; Resources; Supervision; Funding acquisition; Investigation; Project administration; Writing—review and editing Juliette Ravaux: Conceptualization; Data curation; Supervision; Funding acquisition; Methodology; Project administration; Writing—review and editing Nicolas Montagné: Conceptualization; Data curation; Supervision; Investigation; Methodology; Writing—review and editing Thomas Chertemps: Conceptualization; Resources; Data curation; Supervision; Investigation; Methodology; Writing—review and editing Steffen Harzsch: Conceptualization; Resources; Formal analysis; Supervision; Writing—original draft; Writing—review and editing

Funding: Deutsche Forschungsgemeinschaft (DFG): Steffen Harzsch, DFG INST 292/119-1 FUGG; Deutsche Forschungsgemeinschaft (DFG): Steffen Harzsch, DFG INST 292/120-1 FUGG The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.

Data Availability: The raw data of the micro CT scans and the histological section series will be made public at Morph D Base: https://www.morphdbase.de/ We will apply for an accession code immediately. N/A

Ethics: Human Subjects: No Animal Subjects: No

Information for reviewers (full submissions): eLife aims to publish work of the highest scientific standards and importance in all areas of the life and biomedical sciences, from the most basic and theoretical work through to translational, applied and clinical research. Articles must be methodologically and scientifically rigorous, ethically conducted, and objectively presented according to the appropriate community standards. You will be asked for a general assessment and a summary of any major concerns (ideally in fewer than 500 words), as well as a list of any minor comments (optional). You will also have the opportunity to comment on the statistical rigour of the work (optional). In your general assessment, please articulate what is exciting and whether the work represents a significant contribution. Please note our guidelines about requests for additional work: 1. We will only request new work, such as experiments, analyses, or data collection, if the new data are essential to support the major conclusions. The authors must be able to do any new work in a reasonable time frame (additional work should be conducted and written up within two months); otherwise, we will usually reject the manuscript. 2. Any requests for new work must fall within the scope of the current submission and the technical expertise of the authors. Our goal is to make peer review constructive and collaborative: after the reviews have been submitted independently, there is an online discussion between the reviewers in which each reviewer will see the identity of the other reviewers.

eLife Sciences Publications, Ltd is a limited liability non-profit nonstock corporation incorporated in the State of Delaware, USA with company number 5030732, and is registered in the UK with company number FC030576 and branch number BR015634 at the address 24 Hills Road, Cambridge, CB2 1JP. 1 Neuroanatomy of the deep hydrothermal vent shrimp Rimicaris exoculata

2 provides insights into the evolution of higher integrative centers in Crustacea

3

4 Machon J.1, Krieger J.2, Meth R.2, Zbinden M.1, Ravaux J.1, Montagné N.3, Chertemps T.3, 5 Harzsch S.2* 6 7 * corresponding author 8 1 Sorbonne Université, UMR CNRS MNHN 7208 Biologie des organismes et écosystèmes aquatiques 9 (BOREA), Equipe Adaptation aux Milieux Extrêmes, Bât. A, 4e étage, 7 Quai St Bernard, 75005 Paris, 10 France 11 2 University of Greifswald, Zoological Institute and Museum, Department of Cytology and 12 Evolutionary Biology, D-17498 Greifswald, Germany 13 3 Sorbonne Université, UPEC, Univ Paris Diderot, CNRS, INRA, IRD, Institute of Ecology & 14 Environmental Sciences of Paris (iEES-Paris), 4 place Jussieu, 75005 Paris, France 15 16 Abstract 17 18 Alvinocaridid shrimps are emblematic representatives of the deep hydrothermal vent fauna 19 at the Mid-Atlantic Ridge. They are adapted to a mostly aphotic habitat with extreme 20 physicochemical conditions in the vicinity of the hydrothermal fluid emissions. Here, we investigated 21 the brain architecture of the iconic vent shrimp Rimicaris exoculata to understand possible 22 adaptations of its nervous system to the sensory landscape in its extreme habitat. Its brain is 23 modified from the crustacean brain ground pattern by featuring small visual and olfactory neuropils 24 that contrast with well-developed higher integrative centers, the remarkable hemiellipsoid bodies. 25 We propose that these structures in vent shrimps must fulfil functions in addition to higher order 26 sensory processing and suggest a role in place memory. Our study promotes vent shrimps as 27 fascinating models to gain insights into sensory adaptations to peculiar environmental conditions, 28 and the evolutionary transformation of specific brain areas in Crustacea. 29 Abbreviations

30 31 (Numbers 1-X), cell clusters 1-X; 5HT, serotonin immunoreactivity; A1, antenna 1; A1l, lateral 32 flagellum of antenna 1; A1lNv, lateral antenna 1 nerve; A1m, medial flagellum of antenna 1; A1mNv, 33 medial antenna 1 nerve; A2, antenna 2; A2Nv, antenna 2 nerve; AcN, accessory lobe/neuropil; 34 AMPN, anterior medial protocerebral neuropil; AnN, antenna 2 neuropil; ASTir, allatostatin-like 35 immunoreactivity; b, base region of the olfactory glomerulus; bs, branchiostegite; c, cap region of the

36 olfactory glomerulus; CA, cerebral artery; CAL, lateral cerebral artery; CAM, median cerebral artery; 37 CB, central body; DC, deutocerebrum; dR, degenerated rhabdoms; ENv, eye nerve; HN, hemiellipsoid

38 body neuropil; HNcap, hemiellipsoid body cap region; HNcore, hemiellipsoid body core region; IL, 39 intermediate layer; La, lamina; LAN, lateral antenna 1 neuropil; lF, lateral foramen; Lo, lobula; lPC,

40 lateral protocerebrum; maf, myoarterial formation; mafm, myoarterial formation muscles MAN, 41 median antenna 1 neuropil; MAR, Mid-Atlantic Ridge; Me, medulla; mF, medial foramen; mPC, 42 median protocerebrum; MPN, median protocerebral neuropil; OA, ophthalmic artery; Ob, onion 43 bodies; OBNv, organ of Bellonci nerve; oc, oesophageal connectives; ocp, ocular plate; og, olfactory 44 glomerulus; ON, olfactory neuropil; PB, protocerebral bridge; pc, cluster of pigment cells; PMPN, 45 posterior medial protocerebral neuropil; PNT, projection neuron tract; PNTN, projection neuron tract 46 neuropil; PNTCN, projection neuron tract central neuropil; PT, protocerebral tract; R, retina; r, 47 rostrum; sbc, subcap region of the olfactory glomerulus; sc, scaphocerite; SYNir, synapsin

48 immunoreactivity; T, tapetum; T’d, dorsal secondary tendon; T’v, ventral secondary tendon; Ta,

49 anterior tendon; Td, dorsal tendon; TC, tritocerebrum; TM, terminal medulla neuropil; TN, 50 tegumentary neuropil; TNv, tegumentary nerve; V, vessel; VN, visual neuropils 51 Introduction

52 53 The iconic alvinocaridid shrimps were discovered in 1985 during a mission of the deep 54 submersible vehicle ALVIN (Rona et al., 1986) and are now known to be widely distributed 55 representatives of the deep hydrothermal vent fauna along the Mid-Atlantic Ridge (MAR; 56 Desbruyères et al., 2001, 2000; Gebruk et al., 1997; Segonzac et al., 1993). Active vents are dynamic 57 environments, where geothermally heated seawater, the hydrothermal fluid, discharges from 58 chimneys and cracks in the seafloor. At the MAR, vents occur from 850 to 4080 m depth and the pure 59 hydrothermal fluid, which may be up to 350 °C hot, is anoxic, acid, and enriched in potentially toxic 60 minerals and dissolved gases (Charlou et al., 2010, 2002, 2000). Hydrothermal vent habitats, in 61 addition to high hydrostatic pressure and the complete absence of sunlight, are characterized by 62 steep gradients of temperature and concentration of chemicals (Bates et al., 2010; Johnson et al., 63 1988, 1986; Le Bris et al., 2005). Vent organisms are well adapted to these physicochemical 64 conditions, and alvinocaridid shrimps colonize in high abundance the walls of active chimneys, where 65 the hydrothermal fluid mixes with the surrounding cold (4 °C) and oxygenated seawater. Vent 66 ecosystems rely on chemoautotrophic bacteria as primary producers, which convert reduced 67 chemicals through oxidation, thus providing the energy to fix carbon and to produce organic matter 68 that serves as a nutritional basis for primary consumers (Fisher et al., 2007; Jannasch and Mottl, 69 1985; Ponsard et al., 2013; Van Dover, 2000).

70 The shrimp Rimicaris exoculata (Williams and Rona, 1986) is the most intensely studied vent 71 crustacean due to its high abundance at most sites along the MAR and its remarkable lifestyle (Fig. 72 1A,B; Desbruyères et al., 2001; Gebruk et al., 1997; Segonzac et al., 1993; Van Dover et al., 1988). 73 Specimens of R. exoculata are found from 1600 to 4000 m depth (Lunina and Vereshchaka, 2014) and 74 they form massive aggregations in the vicinity of the chimneys, with up to 3000 ind.m-2 (Segonzac et 75 al., 1993). This species is a strict primary consumer, relying on ectosymbiotic bacteria harbored in its 76 enlarged branchial chambers, through a direct nutritional transfer of bacterial carbon products by 77 trans-tegumental absorption (Corbari et al., 2008; Petersen et al., 2010; Ponsard et al., 2013; Zbinden 78 et al., 2004). The associated bacterial metabolic activities include oxidation of sulfide, iron, methane 79 and hydrogen, suggesting that R. exoculata symbionts could have both nutritional and detoxifying 80 roles for the shrimp (Hügler et al., 2011; Zbinden et al., 2008). Hence, this species is strictly 81 dependent on hydrothermal fluid emissions to supplement its symbionts with reduced compounds, 82 and might possess specific sensory abilities for this purpose. Because R. exoculata preferentially lives 83 close to the hydrothermal fluids, the shrimp constantly has to cope with steep temperature gradients 84 ranging approximately from 4 to 40 °C (Cathalot et al., 2018), and its sensory system might be tuned 85 to efficiently probe this dynamic thermal environment.

86 A fundamental question regarding vent shrimp’s environment and lifestyle is how do they 87 detect hydrothermal emissions and further select their microhabitat. Both abiotic and biotic factors 88 are important to determine the animal’s local distribution at hydrothermal vent sites (Le Bris et al., 89 2005; Luther et al., 2001). Several studies showed that R. exoculata possesses a range of 90 morphological, anatomical and physiological adaptations to the hydrothermal environment, related 91 for instance to ectosymbiosis with bacteria (Casanova et al., 1993; Ponsard et al., 2013; Zbinden et 92 al., 2004), respiration in hypoxic conditions (Hourdez and Lallier, 2006; Lallier and Truchot, 1997), or 93 thermal stress (Cottin et al., 2010; Ravaux et al., 2003). However, the sensory mechanisms and 94 adaptations used by the shrimps to perceive their habitat have only been partially investigated (see 95 references below) despite their importance in understanding the lifestyle of vent shrimp species and 96 their long-term evolution.

97 Vision and chemoreception have been proposed to be the major sensory modalities used by 98 vent shrimp to perceive environmental cues (Chamberlain, 2000; Jinks et al., 1998; Pelli and 99 Chamberlain, 1989; Renninger et al., 1995). In vent shrimps, the stalked compound eyes that 100 characterize most malacostracan crustaceans are modified to form enlarged sessile eyes, which in R. 101 exoculata are located underneath the dorsal carapace (Chamberlain, 2000; Gaten et al., 1998; O’Neill 102 et al., 1995; Van Dover et al., 1989). The eyes cannot form images since the ommatidia lack a dioptric 103 apparatus necessary to refract and focus rays of light, but the retina instead consists of 104 hypertrophied rhabdoms and a reflective subjacent layer, structures that maximize the absorption of 105 light. These anatomical features could represent an adaptation to detect very dim light sources. It 106 was suggested that the animals may perceive the black body radiation emitted by the extremely hot 107 fluid which exits the chimney (Chamberlain, 2000; Pelli and Chamberlain, 1989; Van Dover et al., 108 1989). Furthermore, the animal’s antennal appendages respond to sulfide, suggesting that vent 109 shrimps can detect key chemical components of the hydrothermal fluid (Machon et al., 2018; 110 Renninger et al., 1995), but sulfide detection is not restricted to vent shrimps since antennal 111 responses were also recorded from shallow-water Palaemonid shrimp (Machon et al., 2018). From 112 structural descriptions of their antennae 1 and 2 and chemosensory sensilla, it is not clear whether 113 their chemosensory system presents specific adaptations related to the hydrothermal environment 114 (Machon et al., 2018; Zbinden et al., 2017). One specific feature of these organisms is the dense 115 coverage of their antennal appendages by bacterial communities (Zbinden et al., 2018), whose exact 116 role remains unknown. Nevertheless, their occurrence on the sensory organs suggests a functional 117 significance for the shrimp sensory abilities (Zbinden et al., 2018). 118 Crustacean brain structure is best understood in crayfish, crabs, and clawed and spiny 119 lobsters (reviews e.g. Derby and Weissburg, 2014; Harzsch and Krieger, 2018, Schmidt, 2016). We are 120 interested in exploring adaptive changes of crustacean brain structures that have occurred during 121 their evolutionary radiation into particular habitats and their adoption of specific life styles (e.g. 122 Harzsch et al., 2011; Kenning and Harzsch, 2013; Krieger et al., 2015, 2012, 2010; Meth et al., 2017). 123 Differential investment in certain brain neuropils might reflect the sensory landscape which a certain 124 crustacean species typically exploits, so that studying an animal’s brain anatomy may allow for 125 predictions related to its ecology and lifestyle (Sandeman et al., 2014a). For example, in peracarid 126 and remipedian cave crustaceans, the visual neuropils are absent whereas the central olfactory 127 pathway is well developed in Remipedia, highlighting that these blind animals may rely on olfaction 128 as a major sensory modality in their lightless habitat (Fanenbruck et al., 2004; Fanenbruck and 129 Harzsch, 2005; Stegner et al., 2015; Stemme and Harzsch, 2016). In representatives of the genus 130 Penaeus, the olfactory system is moderately developed, while sophisticated antenna 2 neuropils are 131 present, suggesting that the detection of hydrodynamic stimuli is important for these animals (Meth 132 et al., 2017; Sandeman et al., 1993). Hence, comparing the architecture of the sensory centers 133 among divergent crustacean lineages, across wide evolutionary distances and across diverse life 134 styles, can help to understand structural adaptations to specific sensory environments (review in 135 Sandeman et al., 2014a). Studying crustaceans from extreme habitats is particularly informative in 136 this respect (Ramm and Scholtz, 2017; Stegner et al., 2015). However, the structure of the brain in 137 vent shrimps remains poorly understood (Charmantier-Daures and Segonzac, 1998; Gaten et al., 138 1998). Therefore, the present study sets out to provide a detailed description of the architecture of 139 the R. exoculata brain against the background of the extreme conditions that characterize its habitat, 140 and to ultimately discuss its contribution for crustacean brain evolution. 141 Results

142 143 Gross morphology of the cephalothorax 144 145 The wide cephalothorax of Rimicaris exoculata displays large branchiostegites (bs) which 146 surround voluminous gill chambers (Fig. 1C). The animals do not possess eyestalks but bilaterally 147 paired, wing-shaped eyes with a conspicuous, whitish retina that is fused in the anterior region to 148 form the ocular plate (ocp). The lateral parts of the eye extend further dorsally and towards the 149 posterior region of the cephalothorax (Fig. 1C, 2A). The first pair of antennae (A1) is biramous, with 150 two flagella of similar length (Fig. 1C). The second pair of antennae (A2) consists of a basal element, 151 the scaphocerite (sc), and a long uniramous flagellum, slightly wider than those of the antennae 1 152 (Fig. 1C). Micro-CT scans show that the brain is located in the anterior region of the cephalothorax, 153 and receives main sensory afferences from the antenna 1 (A1Nv) and antenna 2 (A2Nv) nerves 154 anteriorly, from the eye nerves (ENv) posterodorsally, from the tegumentary nerves (TNv) laterally, 155 and from the oesophageal connectives (oc) posteriorly (Fig. 1D). 156 [Figure 1] 157 158 Overview of the brain architecture 159 160 Decapod crustacean brains are subdivided into three successive neuromeres: proto-, deuto- 161 and tritocerebrum. In R. exoculata, these regions form a single, medially located mass (i.e. the 162 median brain) (Fig. 1D, 2A). The visual neuropils are closely associated with the lateral 163 protocerebrum (lPC), at a posterodorsal position (Fig. 2A). This arrangement contrasts with other 164 shallow-water carideans and most decapod crustaceans (see e.g. Cronin and Porter, 2008; Meth et 165 al., 2017), in which the lateral protocerebrum is located at some distance of the median brain, in 166 movable eyestalks (Fig. 2B). The deutocerebrum (DC) is associated with the antenna 1 nerves, and 167 the tritocerebrum (TC) is associated with the antenna 2 nerves (Fig. 2). The brain’s neuraxis is not 168 aligned with body axis in R. exoculata, but is bent dorsally so that the protocerebrum is situated 169 posterodorsally to the deutocerebrum (Fig. 2A). 170 Data from micro-CT scans and aligned serial paraffin sections provided a consistent picture of 171 the brain anatomy that we compiled in both three-dimensional reconstructions (Fig. 3A, 4A-D) and a 172 schematic drawing of the R. exoculata brain (Fig. 3B, C). In the following, for simplicity only one brain 173 hemisphere is described, although mirror symmetrical structures are present in the contralateral 174 hemisphere. 175 [Figures 2,3,4] 176 177 Lateral protocerebrum: the visual neuropils 178 179 R. exoculata presents three successive visual neuropils, which are the lamina (La), medulla 180 (Me), and lobula (Lo), from distal to proximal (Fig. 3, 4, 5). The lamina is thin, flattened and elongated 181 dorsally (Fig. 4A,D, 5 B-D). The cell cluster (1) dorsally covers the lamina (Fig. 3C, 4B,D, 5C). 182 Numerous axon bundles from the entire length of the retina (ENv) converge onto the lamina (Fig. 3, 183 4A-D, 5D). The retina consists of photoreceptor organelles, the rhabdoms (dR) (for which the 184 degradation is ascribed to the damaging exposure to intense light during sampling and manipulation 185 of the specimens at the surface), which overlie a white layer of reflecting cells, the tapetum (T), and 186 clusters of pigment cells (pc) (Fig. 5D, Fig. 11A; Nuckley et al., 1996; O’Neill et al., 1995). The medulla 187 is spherical (Fig. 3, 4A-D, 5B,C) and is connected by thin fibers to the lamina (Fig. 5C), the lobula (Fig. 188 5B) and by a dense fibers tract to the terminal medulla (Fig. 5B, white arrowhead). The lobula is 189 slightly larger than the medulla, and is adjacent to the posterior side of the terminal medulla (TM) 190 (Fig. 3, 4A-D,F, 5A, 11A). The merged cell clusters (2) and (3) cannot be clearly separated and cover 191 both the medulla and the lobula (Fig. 3C, 4B,D, 5A-C). 192 [Figure 5] 193 194 Lateral protocerebrum: the hemiellipsoid body and terminal medulla 195 196 The lateral protocerebrum dominates the R. exoculata brain, with the hemiellipsoid body 197 (HN), the terminal medulla (TM), together with the cell clusters (4) and (5) representing about 25 % 198 of the brain volume. The hemiellipsoid body is well defined, with a voluminous, hemispherical cap

199 region (HNcap) located dorsally (Fig. 3, 4, 5D, 6A-F) and displaying synapsin immunoreactivity (SYNir)

200 (Fig. 6G-I). The core region of the hemiellipsoid body (HNcore) is fused posteriorly with the terminal 201 medulla (Fig. 4F, 6B, D,E). The cap and core regions are separated by an arcuate intermediate layer 202 (IL) (Fig. 4E, 6A,B,D,E,G-I) which receives parallel afferent fibers from the terminal medulla anteriorly 203 (namely the HN-TM tract) (Fig. 3A,B, 4F, 6A-C, black asterisks) and a massive bundle of neurites from 204 somata in the cell cluster (5) at the medial side (Fig. 4E, 5D, 6C-E, white asterisks). Some of the 205 intermediate layer fibers display allatostatin-like immunoreactivity (ASTir) near the cap region (Fig. 206 6G). The intermediate layer is devoid of SYNir (Fig. 6 J). The cap region is characterized by synaptic 207 sites forming microglomeruli (Fig. 6H’) and is also innervated by serotonergic neurons (Fig. 6H). The 208 cell cluster (5) is voluminous (Fig. 3C, 4B, 6B-E,I,J) and contains approximately 30,000 cell somata of 209 the so-called globuli cells (Wolff et al., 2017). The hemiellipsoid body receives input from the 210 olfactory neuropils via the projection neuron tract (PNT) in the posterior region (Fig. 3A, 6B, F).

211 The terminal medulla is a large and complex neuropil. Anteriorly, it is shaped like a sphere 212 (Fig. 3, 4A,C,F, 6A-C), and it connects to the intermediate layer of the hemiellipsoid body via the HN- 213 TM tract (Fig. 3A,D, 4F, 6A-C, black asterisks). Posterior to this region, the terminal medulla is large, 214 crossed by seemingly unstructured networks of fibers (Fig. 3A, 4C,F, 6D-F) and displays SYNir (Fig. 215 6I,J). It is innervated by neurites from the cell cluster (4) (Fig. 6D, white arrowhead), and further 216 connects again to the intermediate layer of the hemiellipsoid body via radiating fiber bundles (Fig. 217 6E, black arrowheads). 218 [Figure 6] 219

220 Median protocerebrum 221 222 The median protocerebrum (mPC) comprises two medially fused neuropils, the anterior 223 (AMPN) and posterior (PMPN) medial protocerebral neuropils. The AMPN connects to the terminal 224 medulla of the lateral protocerebrum anteriorly via the protocerebral tract (PT) (Fig.3B, 6C,D) and 225 the PMPN via the posterior protocerebral tract (PPT), the latter containing neurites with strong 226 serotonin-immunoreactivity (5HTir) (Fig. 7C) and seemingly interconnecting the terminal medulla of 227 both hemispheres. Both, the AMPN and PMPN are separated by the unpaired central body neuropil 228 (CB) (Fig. 3), which displays ASTir (Fig. 7B), weak SYNir (Fig. 7A) and strong 5HTir (Fig. 7C). Overall, 229 the median protocerebrum contains many fibers from serotonergic neurons, partly from the cell 230 cluster (x) (which likely refers to the cell clusters (12), (13) and (17) according to Sandeman et al., 231 1992), which define well the elements of the central complex, i.e. the protocerebral bridge (PB) and 232 the central body (CB), and also the posterior region associated to the posterior protocerebral tract. 233 Posteriorly to the central body, fiber bundles of the projection neuron tracts from both hemispheres 234 meet in a region with strong SYNir, that we will call the projection neuron tract central neuropil 235 (PNTCN) (Fig. 7A). 236 [Figure 7] 237 238 Deutocerebrum 239 240 In the deutocerebrum, a paired neuropil with a conspicuous structure is located laterally, the 241 lobe-shaped olfactory neuropil (ON) (Fig. 3, 4A-D, F, 5D, 8A-F). It is composed of approximatively 180 242 wedge-shaped neuropil units, the olfactory glomeruli (og), which are arranged radially around the 243 periphery of a non-synaptic core (Fig. 3B, 4F, 5D, 8A-F). Each glomerulus shows strong SYNir (Fig. 8E- 244 F), as well as ASTir which highlights a subdivision of each glomerulus into a cap (c), subcap (sbc) and 245 base (b) region (Fig. 8E’). The sensory input of the olfactory neuropil comes from the olfactory 246 sensory neurons innervating the aesthetasc sensilla on the lateral flagellum of the antenna 1. The 247 somata of olfactory interneurons located in the cell cluster (9/11) innervate fibers to the olfactory 248 neuropil, some of which display ASTir (Fig. 8H). These fibers enter via the medial foramen (mF) into 249 the core of the neuropil (Fig. 8C,E), from where they target the glomerular base region (Fig. 8E), or 250 cross to the lateral foramen (lF) (Fig. 8A,C,E) to spread out laterally and innervate the glomerular cap 251 region (Fig. 8E, white arrowhead). The medial foramen is also the place where efferent fibers exit 252 from the olfactory neuropil. These are the axons of the olfactory projection neurons that form the 253 projection neuron tract (Fig. 8B,D). A projection neuron tract neuropil (PNTN) as known from other 254 decapods (e.g. Sandeman et al. 1992, Harzsch and Hansson 2008, Krieger et al. 2012) is identifiable 255 close to the ascending branch of the tract (Fig. 8E,F). The projection neuron tract then transverses 256 the median protocerebrum and projects to the lateral protocerebrum (see above).

257 The lateral antenna 1 neuropil (LAN) is located medially to the olfactory neuropil. It is U- 258 shaped (Fig. 3, 8A,B,H) and displays strong SYNir, as well as ASTir, which reveals a transversely 259 stratified pattern (Fig. 8G). This neuropil connects posterodorsally to the median protocerebrum (Fig. 260 8G,H).

261 The median antenna 1 neuropil (MAN) is small, poorly defined, and located in the center of 262 the deutocerebrum, below the anterior region of the median protocerebrum and between the paired 263 lateral antenna 1 neuropils (Fig. 3A,B). 264 [Figure 8] 265 266 Tritocerebrum 267 268 The tritocerebrum comprises the antenna 2 neuropil (AnN), which has a cylindrical shape and 269 lies in front of the oesophageal connectives (Fig. 3, 4, 8HA). SYNir and ASTir show a transversely 270 stratified pattern within this neuropil (Fig. 8I). Poorly differentiated from the antenna 2 neuropil, the 271 tegumentary neuropil (TN) is located posterodorsally (Fig. 3). 272 273 The organ of Bellonci 274 275 The organ of Bellonci (OB) is typical for many crustaceans but its sensory function remains 276 unclear (Chaigneau, 1994). In R. exoculata, this organ is conspicuous and comprises onion bodies 277 (Ob) structures connected to a well-developed nerve tract (OBNv). The onion bodies are situated on 278 the anterolateral side of the brain, in front of the hemiellipsoid body (Fig. 3, 4A-D, 5D, 9A). They 279 represent a cluster of about fifty densely packed lobules (Fig. 9A,B), many of them containing 280 elements of granular appearance (Fig. 9B’, white arrowhead). Some lobules are further located in the 281 proximal region of OBNv (Fig. 9B). This nerve is large in its proximal region, and progressively tapers 282 as it draws away from the brain (Fig. 9A). Anterodorsally, the nerve extends through the retinal 283 layers and connects underneath the cuticle of the ocular plate (Fig. 9C,C’). 284 [Figure 9] 285 286 The myoarterial formation and cerebral vascular system 287 288 The myoarterial formation (maf) (or cor frontale, auxiliary heart) underlies the dorsal 289 carapace and is located between the paired eyes, above the brain (Fig. 10A,B,C). This organ is 290 voluminous, being almost as long as the elongated retina, and extends ventrally towards the dorsal

291 region of the brain (Fig. 10B’,C’). Two adjacent and parallel muscle bundles (mafm) penetrate through 292 the myoarterial formation (Fig. 10A) and attach to the cuticle via tendons located either anteriorly

293 (Ta) or dorsally (Td) (Fig. 10B,C). Two thinner muscular bundles cross the myoarterial formation in its 294 middle region, perpendicular to the main adjacent muscles, and are attached both to the dorsal and

295 ventral cuticle of the cephalothorax, by secondary dorsal and ventral tendons (T’d, Fig. 10B,C; T’v, Fig. 296 10B’).

297 Anteriorly, below the junction of the ocular plate with the dorsal carapace, the myoarterial 298 formation gives rise to three conspicuous, large cerebral arteries, a central cerebral artery (CA) and 299 two ophthalmic arteries (OA), which all make a steep U-turn and extend parallel towards the dorsal 300 side of the brain (Fig. 10B,C, arrowhead). In an anterodorsal position, median to the hemiellipsoid

301 bodies, the central cerebral artery divides into three smaller arteries, one median (CAM) and two

302 lateral ones (CAL) (Fig. 10B’,C’). The median artery passes over the brain, between the two spherical 303 masses of the lateral protocerebrum, and then divides into two branches, one entering the brain 304 posteriorly, at the level of the median protocerebrum, and a larger one merging with the ventral 305 region of the myoarterial formation. This suggests a loop system wherein part of the hemolymph in 306 the cerebral artery goes back into the myoarterial system. The lateral cerebral arteries are coil- 307 shaped and enter the brain above the insertion of the antenna 1 nerve to target for instance the 308 olfactory neuropils and the lateral antenna 1 neuropils in the deutocerebrum. The ophthalmic 309 arteries enter the brain in a posterodorsal position (Fig. 10C’) and target the visual neuropils and the 310 lateral protocerebrum. 311 Anteriorly, below the junction of the ocular plate with the dorsal carapace, the myoarterial 312 formation gives rise to three conspicuous, large cerebral arteries, a central cerebral artery (CA) and 313 two ophthalmic arteries (OA), which all make a steep U-turn and extend parallel towards the dorsal 314 side of the brain (Fig. 10B,C, arrowhead). In an anterodorsal position, median to the hemiellipsoid

315 bodies, the central cerebral artery divides into three smaller arteries, one median (CAM) and two

316 lateral ones (CAL) (Fig. 10B’,C’). The median artery passes over the brain, between the two spherical 317 masses of the lateral protocerebrum, and then divides into two branches, one entering the brain 318 posteriorly, at the level of the median protocerebrum, and a larger one merging with the ventral 319 region of the myoarterial formation. This suggests a loop system wherein part of the hemolymph in 320 the cerebral artery goes back into the myoarterial system. The lateral cerebral arteries are coil- 321 shaped and enter the brain above the insertion of the antenna 1 nerve to target for instance the 322 olfactory neuropils and the lateral antenna 1 neuropils in the deutocerebrum. The ophthalmic 323 arteries enter the brain in a posterodorsal position (Fig. 10C’) and target the visual neuropils and the 324 lateral protocerebrum.

325 The cerebral vascular system of R. exoculata is considerably developed, with blood vessels 326 supplying all brain neuropils and cell clusters, as in other crustaceans, including large vessels that 327 irrigate the visual neuropils (Fig. 5A,C), the deutocerebrum (Fig. 8B) and the lateral protocerebrum 328 (Fig. 9B,B’). The Azan staining reveals pink-to-purple cerebral arteries that enter the brain (CA) (Fig. 329 5C,D, 8A,B, 10A), and orange vessels inside the brain (V) (Fig. 5A,C, 8B, 9B). 330 [Figure 10] 331 332 Comparisons of the olfactory system and the higher integrative centers with other crustaceans

333 334 Table 1 presents a comparison of aesthetasc and olfactory neuropil characteristics in 335 different taxa of crustaceans. The number of olfactory glomeruli in R. exoculata fits within the range 336 displayed by other decapods, but their unitary volume is in the lower range, leading to relatively 337 small olfactory neuropils (excluding the fibrous core) compared to other species (Table 1, Fig. 11e). In 338 contrast, the higher integrative centers (i.e. the hemiellipsoid body and the medulla terminalis, 339 associated to cell cluster 5) are especially well-developed in R. exoculata compared to other 340 crustaceans and in relation to the relative size of the olfactory neuropils (Fig. 11). As an example, 341 from relative volumes obtained from 3D reconstructions, the higher integrative centers in R. 342 exoculata occupy approximatively 25 % of the total brain volume, similarly to the caridean shrimp 343 Palaemon elegans (22 %) but twice more than in the giant robber crab Birgus latro (13 %), whereas 344 their olfactory neuropils are respectively almost twice (4.2 %) and six (17 %) times more voluminous 345 than those of R. exoculata (2.7 %). 346 [Table 1] 347 [Figure 11]

348 349 Discussion

350 351 General remarks 352 353 The emblematic alvinocaridid shrimp Rimicaris exoculata is an endemic species to 354 hydrothermal vent habitats, well adapted to these deep sea environments with peculiar 355 physicochemical conditions. The present study sets out to gain insights into adaptations to specific 356 features of the vent habitat (e.g. low ambient light levels and steep variations of chemical 357 concentrations). The analysis of the brain architecture in R. exoculata aims to highlight relative 358 investments into certain neuronal subsystems, in relation with the animal’s habitat and lifestyle. The 359 general anatomy of the brain of R. exoculata corresponds in many aspects to the ground pattern of 360 the malacostracan crustacean brain (Kenning et al. 2013), including the subdivision into proto-, 361 deuto- and tritocerebrum, the location of main nerves and the presence of distinct cell clusters. 362 However, the brain of R. exoculata also exhibits morphological differences to other malacostracans, 363 especially at the level of the lateral protocerebrum. The organ of Bellonci is especially conspicuous 364 (also observed by Charmantier-Daures and Segonzac, 1998), but its sensory function remains elusive. 365 In the following, we will focus on the structure of major sensory centers (i.e. the visual system, the 366 olfactory system and the higher integrative centers). We will also discuss the evolution of the 367 hemiellipsoid bodies as higher integrative brain centers, which are remarkable in R. exoculata. We 368 will begin our account by addressing the neurovascular system that supplies the brain. 369 370 The neurovascular system 371 372 In crustaceans, the neurovascular system has been described mainly in crayfish (Chaves da 373 Silva et al., 2012; Scholz et al., 2018), crabs (McGaw, 2005; McGaw and Reiber, 2002; Sandeman and 374 Callan 1967) and spiny lobsters (Steinacker, 1979) (reviews in McMahon, 2001; Steinacker, 1979, 375 1978; Wilkens, 1999). The brain and eyes are supplied in hemolymph via the anterior aorta system, 376 which originates antero-medially from the heart and runs between the stomach and the dorsal 377 integument (Scholz et al., 2018). Anteriorly, a dilatation of the anterior aorta, the myoarterial 378 formation (Scholz et al., 2018; also named the cor frontale in e.g. McGaw, 2005; Steinacker, 1978) 379 which functions as an auxiliary heart, pumps the hemolymph specifically towards the anterior part of 380 the central nervous system. In malacostracan crustaceans, the myoarterial formation above the brain 381 gives rise to a descending cerebral artery, which supplies the median brain, and to two ophthalmic 382 arteries that turn laterally and extend into the eyestalks to supply the visual neuropils (Chaves da 383 Silva et al., 2012; McGaw, 2005; Scholz et al., 2018). In R. exoculata, consistent with the absence of 384 eyestalks, the myoarterial formation and its arteries differ in shape, size and position from those 385 previously described in other malacostracans. Among the potential corollaries for the pronounced 386 neurovascular system in R. exoculata, one is the more efficient hemolymph pumping to the brain. In 387 crustaceans, the perfusion of the brain is modulated by physiological or environmental factors, such 388 as hypoxia (Reiber and McMahon, 1998). Because the pure hydrothermal fluid is anoxic, the mixing 389 of the fluid with the surrounding seawater can create hypoxic conditions for vent animals (Childress 390 and Fisher, 1992; Schmidt et al., 2008). Known adaptations to hypoxia in vent crustaceans include an 391 hemocyanin with a higher affinity for oxygen compared to shallow-water species (Chausson et al., 392 2004; Lallier and Truchot, 1997; Sanders et al., 1988). A very pronounced capillary network was also 393 observed in hydrothermal vent alvinellid polychaetes (Hourdez and Lallier, 2006). Accordingly, the 394 pronounced myoarterial formation and large cerebral arteries in R. exoculata could represent a 395 particularly efficient system for oxygen delivery to the brain to cope with low availability of oxygen. 396 397 A visual system adapted to a dim light environment 398 399 Van Dover and co-workers (1989) described the R. exoculata eyes as a pair of large anteriorly 400 fused organs that underlie the transparent dorsal carapace of the cephalothorax and demonstrated 401 the presence of rhodopsin-like visual pigments in high quantity, with a maximum absorption at 500 402 nm. Subsequent analyses showed that the eyes comprise a smooth cornea located above a dense 403 layer of hypertrophied rhabdoms, under which a white layer of reflective cells, the tapetum, is 404 located and maximizes the absorption of light by the photoreceptors (Chamberlain, 2000; Jinks et al., 405 1998; Nuckley et al., 1996; O’Neill et al., 1995). These elements of the retina were all discernible in 406 our histological sections, although the rhabdoms were strongly degenerated, a process ascribed to 407 the damaging exposure to intense light during sampling (Herring et al., 1999; Johnson et al., 1995). 408 The eyes of R. exoculata lack the dioptric apparatus which characterizes the ommatidia of 409 compounds eye pelagic and shallow water crustaceans and thus cannot form images (Chamberlain, 410 2000; Jinks et al., 1998; Nuckley et al., 1996; O’Neill et al., 1995), but their highly sensitive naked 411 retina seems adapted for the detection of low ambient light levels, to the detriment of spatial 412 resolution (Chamberlain, 2000; Van Dover et al., 1989).

413 In malacostracan crustaceans, the visual input from the compound eyes is processed by a 414 suite of retinotopic visual neuropils, usually but not exclusively located within the moveable 415 eyestalks (Fig. 2B) (Strausfeld, 2012, Loesel et al., 2013). The absence of eyestalks of R. exoculata 416 coincides with a strong size reduction and fusion of the visual neuropils with the median brain (Fig. 417 2A). Nevertheless, already Charmantier-Daures and Segonzac (1998) and Gaten and co-workers 418 (1998) differentiated three visual neuropils in R. exoculata, namely the lamina, medulla, and lobula, 419 as in the ground pattern of the Malacostraca. However, in R. exoculata these neuropils are located 420 posterodorsally to the enlarged lateral protocerebrum. The dorsal expansion of the flattened lamina, 421 that extends in parallel to the retina, suggests a retinotopic projection of photoreceptor input onto 422 the lamina which may allow the animals to extract directional information from light sources above. 423 In the medulla, immunohistochemistry revealed an outer layer (which is also faintly visible in 424 histological sections), suggesting a subdivision of the medulla into an outer and inner region, as seen 425 in crayfish (Strausfeld and Nässel, 1981). No such stratification was observed for the lobula, and 426 synapsin immunoreactivity was weak in this most proximal neuropil, although in malacostracans with 427 well-developed compound eyes, the lobula displays numerous, neurochemically diverse strata (e.g. 428 Brachyura and Anomura, Harzsch and Hansson, 2008; Krieger et al., 2012, 2010; Wolff et al., 2012; 429 Astacidea, Polanska et al., 2007; Dendrobranchiata, Meth et al., 2017; Stomatopoda, Strausfeld, 430 2005). The simplified structure of the lobula which in other malacostracans plays a role in motion 431 detection (Strausfeld, 2012) may mirror the inability of the eye to form images. Also, the lobula plate, 432 a fourth visual neuropil present in several malacostracan taxa (e.g. Brachyura and Anomura, 433 Bengochea et al., 2017; Harzsch and Hansson, 2008; Krieger et al., 2012, 2010; Sztarker et al., 2009; 434 Dendrobranchiata, Meth et al., 2017; Stomatopoda, Strausfeld, 2005; Leptostraca, Kenning et al., 435 2013; Isopoda, Kenning and Harzsch, 2013; Sinakevitch et al., 2003) could not be identified in R. 436 exoculata. The lobula plate has been suggested to mediate optokinetic control, necessary to track 437 moving objects (e.g. conspecifics, preys, predators) (Sztarker et al., 2005). Such a role is consistent 438 with the loss of the lobula plate in R. exoculata, which lacks the realization of image formation, 439 necessary for tracking moving objects.

440 Many eyeless representatives of Crustacea have partially or totally lost their central visual 441 pathways (e.g. Peracarida, Ramm and Scholtz, 2017; Stegner et al., 2015; Cephalocarida, Elofsson and 442 Hessler, 1990; Stegner and Richter, 2011; Remipedia, Fanenbruck et al., 2004; Mystacocarida, 443 Brenneis and Richter, 2010), potentially under the selective pressure that favors a reduction of these 444 nervous tissues to limit the amount of energy expended on their function (Klaus et al., 2013; Moran 445 et al., 2015; Niven and Laughlin, 2008). Contrary, the fact that neuronal elements indicative for a 446 functional visual system are present in R. exoculata must mean that there is light to exploit as an 447 environmental cue. Also, the unusual nature of the visual system of R. exoculata suggests that it 448 exploits a specific type of signal. One prominent hypothesis refers to the thermal black body 449 radiation emitted by the hot hydrothermal fluid at the chimney’s exit with a temperature of up to 450 350 °C, which peaks in the infrared but part of its spectrum extends into the visible light (Pelli and 451 Chamberlain, 1989; Van Dover et al., 1996, 1988; Van Dover and Fry, 1994). The ability to localize this 452 radiation could serve both to attract the shrimp to optimal areas for supplying its symbionts with 453 vital, reduced compounds of the hydrothermal fluid, and to allow avoidance of scorching fluid (Van 454 Dover et al., 1989). Visual cues other than thermal radiation are likely to be also exploited by R. 455 exoculata, related to turbulence, mixing and precipitation, such as chemi-, crystallo-, tribo- and sono- 456 luminescence, for which the emission spectra lie between 450-800 nm (Reynolds and Lutz, 2001; 457 Tapley et al., 1999; Van Dover et al., 1996; Van Dover and Fry, 1994; White, 2000; White et al., 2002). 458 459 The olfactory system 460 461 Two modes of chemoreception, linked to distinct chemosensory pathways, are distinguished 462 in malacostracan crustaceans (Derby and Weissburg, 2014; Schmidt and Mellon, 2010): olfaction, 463 which is mediated by the aesthetasc sensilla located on the lateral flagellum of the antenna 1, and 464 distributed chemoreception, which is mediated by the bimodal chemo- and mechanosensory sensilla 465 located mainly on all antennal appendages, the mouthparts, and the walking appendages (Garm et 466 al., 2005, 2003; Garm and Watling, 2013; Mellon, 2014, 2012; Schmidt and Gnatzy, 1984). R. 467 exoculata presents aesthetascs in similar number and dimensions to other caridean representatives 468 (Table 1), as well as several bimodal sensilla with different morphologies on the antennal 469 appendages (Zbinden et al., 2017).

470 Olfaction has been extensively studied in malacostracans (e.g. Ache, 2002; Derby and 471 Weissburg, 2014; Schmidt and Mellon, 2010), and the central olfactory pathway has received much 472 attention in crustacean neuroanatomy (e.g. Blaustein et al., 1988; Harzsch and Krieger, 2018; 473 Kenning et al., 2013; Kenning and Harzsch, 2013; Krieger et al., 2015, 2012, 2010; Sandeman et al., 474 1992; Schachtner et al., 2005; Schmidt and Mellon, 2010). The afferent olfactory input from the 475 olfactory sensory neurons innervating the aesthetascs targets the conspicuous olfactory neuropils, 476 which are lobe-shaped and bilaterally arranged in the deutocerebrum (Fig. 8A-F). They are composed 477 of spherical or cone-shaped dense synaptic neuropils, namely the olfactory glomeruli, which are 478 radially arranged around the periphery of a core of non-synaptic fibers. The olfactory glomeruli are 479 subdivided into a cap, subcap and base regions in several decapod taxa (e.g. Harzsch and Krieger, 480 2018; Schachtner et al., 2005; Schmidt and Ache, 1997). The glomeruli of R. exoculata appear to 481 conform to this principle design, with an identical subdivision (Fig. 8E’). Although the number of 482 olfactory glomeruli is in the same range to that of its close relative Palaemon elegans, the olfactory 483 neuropils of R. exoculata are relatively small in terms of volume and not overly developed compared 484 to other species (Table 1 and Fig. 11). Hence, the dimensions and structural complexity of the 485 olfactory neuropils in R. exoculata do not suggest, judging from comparative brain anatomy, that the 486 loss of the eye’s capacity to form images is compensated by sophisticated olfactory abilities.

487 Efficient olfactory abilities would have been especially relevant to probe the chemical 488 environment of R. exoculata, which is dynamic, with strong concentration variations of hydrothermal 489 fluid chemicals as the hydrothermal fluid dilutes with the surrounding seawater. Sulfide and other 490 chemicals could serve as highly important environmental cues for R. exoculata (Renninger et al. 491 1995, Machon et al. 2018) to locate active edifices as optimal areas to supply its chemoautotrophic 492 symbionts with reduced compounds. However, sulfide detection is likely mediated by distributed 493 chemoreception, or both distributed chemoreception and olfaction, rather than exclusively olfaction, 494 as it can detected by the flagella of the antenna 2 which does not bear aesthetascs (Machon et al., 495 2018). Olfaction is also involved in the recognition of conspecifics (Thiel and Breithaupt, 2011) and 496 the localization of sexual partners (Wyatt, 2014), but there is to date no detailed information on the 497 inter-individual interactions in and out of the swarms of R. exoculata. The detection of chemical cues 498 produced by bacteria could also appear especially relevant since the sensory antennal appendages of 499 vent shrimp are often covered by a dense bacterial layer, whose roles are unknown yet (Zbinden et 500 al. 2018). 501 502 Evolution of higher integrative brain centers: the hemiellipsoid body 503 504 Malacostracan crustaceans display a rich repertoire of complex behavioral patterns related 505 to finding food, shelter and mating partners, kin recognition and brood care, as well as orientation 506 and homing. Decapod crustaceans are also known for complex social interactions such as communal 507 defensive tactics, the occupation of common shelters, cooperative behavior during long-distance, 508 offshore seasonal migration and the establishment of dominance hierarchies (Breithaupt and Thiel, 509 2011; Derby and Thiel, 2014; Duffy and Thiel, 2007; Thiel and Watling, 2015). Because such complex 510 behaviors most likely involve elements of learning and memory, higher integrative brain centers are 511 suggested to provide the neuronal substrate for more sophisticated processing underlying such 512 behaviors (review in Sandeman et al., 2014a). Such centers receive input exclusively from second or 513 higher order neurons but not from any primary sensory afferents (i.e. from the peripheral nervous 514 system) and contain interneurons responding to the stimulation of several different sensory systems. 515 In the malacostracan brain, the (bilaterally paired) terminal medulla, hemiellipsoid body, and 516 accessory lobe seem to function as higher integrative centers, all three distinct neuropil areas which 517 display a high level of complexity and are notable for their substantial volume (Sandeman et al., 518 2014a). The terminal medulla and the closely associated hemiellipsoid body, are targeted by axons 519 of the olfactory projection neurons as output pathway of the olfactory neuropil and accessory lobe 520 (where present; reviews Derby and Weissburg, 2014; Harzsch and Krieger, 2018; Schmidt, 2016). 521 Because of these anatomical relation, evolutionary (Sullivan and Beltz, 2004, 2001) and functional 522 considerations (Harzsch and Krieger, 2018; Sandeman et al., 2014a; Strausfeld, 2012) have focused 523 on possible roles of these centers in higher order olfactory processing. In addition to the olfactory 524 projection neuron axons, the terminal medulla also receives input from the visual neuropils in several 525 malacostracans (reviewed in Sandeman et al., 2014a). A specific type of local interneurons associated 526 to the medulla terminalis and the hemiellipsoid body are the parasol cells (Mellon and Alones, 1997; 527 McKinzie et al., 2003; Mellon et al., 1992; Mellon, 2000; Mellon et al., 1992) which respond to 528 olfactory, tactile, and visual stimuli, thus highlighting their role as elements in higher order 529 integration (Mellon and Alones, 1997; Mellon, 2003, 2000; Mellon and Wheeler, 1999). Recent 530 evidence obtained from a brachyuran crab suggests an involvement of the crustacean hemiellipsoid 531 body/terminal medulla complex in memory processes (Maza et al., 2016). Furthermore, considering 532 anatomical similarities of the crustacean hemiellipsoid body and insect mushroom body, Wolff and 533 co-workers (2017) suggested an involvement in place memory. 534 [Figure 12] 535 536 During the evolutionary elaboration of malacostracan brains, substantial modifications 537 occurred related to the relative proportion of secondary sensory input and investment in size of the 538 various higher integrative centers (Fig. 11; Harzsch and Krieger, 2018; Sandeman et al., 2014a). 539 Because the terminal medulla has a highly complex and highly variable structure, being composed of 540 several, partly confluent neuropil lobes with heterogeneous appearance containing both coarse and 541 fine fibers (e.g. Blaustein et al., 1988), its architecture so far has not been studied in a comparative 542 context. We will focus in the following on the hemiellipsoid body whose structure is somewhat easier 543 to grasp (Fig. 11). In its simplest form, the hemiellipsoid body consists of a volume of fine neuropil 544 with little texture that is closely associated with the terminal medulla (Fig. 11, 12). Such a phenotype 545 is for example common in leptostracans, the presumably most basal branch of the Malacostraca 546 (Kenning et al., 2013), but also in representatives of the Dendrobranchiata (Meth et al., 2017; 547 Sullivan and Beltz, 2004), and several Brachyura (Krieger et al., 2010, 2012b, 2015). Isopoda as 548 representatives of the Peracarida also feature simple, dome-shaped hemiellipsoid bodies (Fig. 11, 12; 549 Kenning and Harzsch, 2013; Stemme and Harzsch, 2016) whereas in Amphipoda (Ramm and Scholtz, 550 2017) and blind groups of peracarids from relict habitats (Stegner et al., 2015), this center is poorly 551 developed and may be entirely missing. A more complex phenotype features a separation of the 552 hemiellipsoid body into two separated areas, an architecture present for example in the spiny 553 lobsters (neuropils I and II, Blaustein et al., 1988), the crayfish Procambarus clarkii and Orconectes 554 rusticus (neuropil I and II, Sullivan and Beltz, 2001), and Cherax destructor (HBI and HBII, Sullivan and 555 Beltz, 2005). The clawed lobster Homarus americanus also features two neuropil units, but these are 556 stacked on top of each other as cap and core neuropils separated by an intermediate, non-synaptic 557 layer (Sullivan and Beltz, 2001). Additional differences exist between the crayfish and the clawed 558 lobster concerning the areas that are targeted by the axons of the projection neurons. (Mellon et al., 559 1992; Mellon et al., 1992; Sullivan and Beltz, 2005, 2001). Hemiellipsoid bodies with a cap/core 560 structure separated by an intermediate layer are also present in the brains of marine (Krieger et al., 561 2012b) and terrestrial hermit crabs of the taxon Coenobitidae, Coenobita clypeatus (Harzsch and 562 Hansson, 2008; Polanska et al., 2012; Wolff et al., 2012) and Birgus latro (Krieger et al., 2010). These 563 animals all display a large hemiellipsoid body with a peripheral, dome-shaped cap neuropil enclosing 564 two dome-shaped core neuropil areas Core 1 and Core 2 (Fig. 10, 11). Their hemiellipsoid body is 565 associated with several thousands of small, intrinsic neurons (Harzsch and Hansson, 2008; Krieger et 566 al., 2010) The cap and core neuropils are separated by intermediate layers formed by the neurites of 567 these intrinsic interneurons and the afferents of the projection neuron tract in a rectilinear 568 arrangement (Wolff et al., 2012). 569 In the hemiellipsoid bodies of the stomatopod crustaceans Gonodactylus bredenii (Sullivan 570 and Beltz, 2004) and Neogonodactylus oerstedii (Wolff et al., 2017), the cap/core motif is modified 571 such that the cap layer (termed “calyx” in Wolff et al., 2017) is much thinner than the core neuropil 572 and that the cluster of intrinsic neurons expands over much of the surface of the cap neuropil. The 573 additional stalked neuropils in the lateral protocerebrum of N. oerstedii (Wolff et al., 2017) will not 574 be discussed here for simplicity. In Stenopus hispidus (Stenopodidea), the hemiellipsoid body appears 575 very complex in structure, with apparently three distinct lobular neuropils (Sullivan and Beltz 2004; 576 Krieger et al., unpublished). The hemiellipsoid body in the caridean species P. elegans and 577 Palaemonetes pugio also presents three lobular neuropils, two of which present a cap layer and one 578 or two core regions, and a third neuropil without clear subdivision (Fig. 11D, Fig. 12; Sullivan and 579 Beltz 2004). The hemiellipsoid body of R. exoculata in many aspects, closely corresponds to the 580 cap/core layout although it is slightly simpler than in Coenobitidae with only one core neuropil, 581 similar to the arrangement observed in H. americanus (Sullivan and Beltz, 2001) (Fig. 12).

582 In summary, the hemiellipsoid body displays more structural variations across the 583 Malacostraca than many other elements of the crustacean brain areas which led Sandeman and co- 584 workers (2014) to note that, within the Malacostraca, several different evolutionary trajectories are 585 present to increase their brain’s capacity for integrating olfactory and multimodal stimuli. This 586 diversity masks common motifs of hemiellipsoid body architecture, explaining why genealogical 587 relationships of the crustacean and insect protocerebral multimodal centers have been discussed 588 controversially for many years (reviews e.g. Loesel et al., 2013; Sandeman et al., 2014b; Strausfeld, 589 2012, 2009, 1998). Latest evidence suggests that, despite many morphological differences, these 590 protocerebral structures of insects and crustaceans nevertheless share common architectural, 591 physiological and neurochemical features suggesting a homology of their very basic neuronal 592 circuitry (Brown and Wolff, 2012; Maza et al., 2016; Wolff et al., 2012, 2017; Wolff and Strausfeld, 593 2015). 594 595 Possible functions of the hemiellipsoid body: new lesson from R. exoculata? 596 597 Because the projection neuron tract provides a massive input to the lateral protocerebrum, 598 recent comparative considerations have suggested that the structural elaboration and size of 599 hemiellipsoid bodies largely mirror the importance of the central olfactory pathway in a given brain, 600 thus emphasizing their role in higher order olfactory processing (e. g. Harzsch and Krieger, 2018; 601 Sandeman et al., 2014). Along these lines, Harzsch and Hansson (2008) and Krieger and co-workers

602 (2010) noted that in representatives of the Coenobitidae, the architectural complexity and volume of 603 the olfactory neuropil closely correlates to that of the hemiellipsoid body. The comparative plates 604 (Fig. 11 and 12) demonstrate that R. exoculata dramatically deviates from this pattern in that their 605 disproportionally large hemiellipsoid body contrasts with an inconspicuous, small olfactory neuropils. 606 The observation that visual input is likely to also play a subordinate role in these animals compared 607 to shallow-water relatives with fully developed compound eyes makes us suggest that in R. 608 exoculata, their impressive hemiellipsoid body must fulfil additional functions in addition of higher 609 order sensory processing. Discussing anatomical similarities of the crustacean hemiellipsoid body and 610 insect mushroom body, Wolff and co-workers (2017a) suggested for these two neuropils a role in 611 place memory, based on observations that insects with elaborate navigational skills display elaborate 612 mushroom bodies. Considering recent experiments that suggest an involvement of the crustacean 613 hemiellipsoid body/terminal medulla-complex in memory processes (Maza et al., 2016), we here 614 propose that the hemiellipsoid body in R. exoculata is involved in the formation of place memory. 615 This hypothesis is further supported by the presence of serotonergic tracts within the hemiellipsoid 616 bodies (Fig. 6H), since serotonin has a function for place memory and learning in the mushroom 617 bodies of Drosophila melanogaster (e.g. Sitaraman et al., 2008). Spiny lobsters Panulirus argus are 618 renowned for their extensive offshore migrations and their ability to orient accurately towards their 619 home sites over long distances by using the direction of water movement (surge) caused by wave 620 action, learned local structural features, and geomagnetic cues for navigation (reviewed in Sandeman 621 et al., 2014). Using a GPS-based telemetric system, giant robber crabs, Birgus latro, were shown to 622 form route memories and may use path integration as navigation strategy and in translocation 623 experiments were shown to be capable of homing over large distances (Krieger et al., 2012a). The 624 above mentioned crustacean species display impressive hemiellipsoid bodies. For survival in the 625 extreme, lightless habitat of R. exoculata, an excellent place memory may be essential for avoiding 626 the dangerously hot vent chimneys and memorizing emission sites of hydrothermal fluids rich in 627 those chemicals on which their endosymbiont bacteria depend. 628 629 Conclusion 630 631 Our observations of the general brain architecture of R. exoculata highlight several specific 632 peculiarities, which could be related to adaptations to the specific sensory landscape of the vent 633 habitat. The well-developed neurovascular system could be particularly efficient for brain 634 oxygenation, to cope with the low availability of oxygen in the close surroundings of active chimneys. 635 The conservation of the visual pathway and neuropils in a mostly aphotic environment suggests that 636 vision nevertheless is a relevant sense for vent shrimp, although they lack effective motion detection 637 abilities. The olfactory system does not present specific traits and is probably not a dominant sensory 638 modality in this shrimp unlike what has been proposed so far. On the other hand, the higher 639 integrative centers are especially well-developed, and the disproportionally large hemiellipsoid 640 bodies could be involved in complex integrative processes such as place memory, which could be a 641 primary navigational cue. Overall, vent shrimp appear to be especially interesting models to 642 investigate both sensory adaptations to peculiar environmental conditions, and the evolution of the 643 sensory centers among Crustacea. 644 Materials and Methods

645 646 Animal collection and fixation procedures 647 648 Specimens of alvinocaridid shrimp R. exoculata (Williams and Rona, 1986) were collected on 649 the TAG vent site (MAR, 26°08’N-44°49’W, 3600 m depth) during the BICOSE 2018 cruise on the 650 Research Vessel ‘Pourquoi Pas?’. Animals were sampled with the suction device of the Diving Support 651 Vessel ‘Nautile 6000’, and recovered at their in situ pressure using the PERISCOP isobaric recovery 652 device (Shillito et al., 2008). Immediately after retrieval, specimens were dissected to remove the 653 hepatopancreas prior to fixation. The specimens for histology and x-ray micro-computed tomography 654 (micro-CT) scans were stored in Bouin’s fixative (10 % formaldehyde, 5 % glacial acetic acid in 655 saturated aqueous picrinic acid) at 4 °C until use. The specimens for immunohistochemistry were 656 fixed 24 to 48 h in 4 % formaldehyde (FA) in 0.1 M Phosphate buffered saline (PBS) at 4 °C for 24 h,

657 and then stored in 0.1 M PBS with NaN3 at 4 °C until use. All specimens were sexed using the sexual 658 dimorphism from the second pair of pleopods. Specimens are females for all micro-CT and histology 659 experiments, and are both females and males for immunohistochemistry. 660 Caridean shallow water shrimp Palaemon elegans (Rathke, 1837) were collected from Saint- 661 Malo Bay (France; 48°64’N,-2°00’W), in January 2018, using a shrimp hand net. Specimens were 662 dissected and fixed as described above. Protocols for other species are described in the following: 663 Nebalia herbstii, Kenning et al., 2013; Penaeus vannamei, Meth et al., 2017; Saduria entomon, 664 Kenning and Harzsch, 2013; Carcinus maenas, Krieger et al., 2012; Birgus latro, Krieger et al., 2010. 665 666 Histology 667 668 The heads of Bouin-fixed animals were dehydrated in a graded series of ethanol and 669 embedded in paraffin wax mixed with 5 % beeswax. Serial sections (7 µm) were taken in the frontal 670 or sagittal plane with a microtome (Leica RM 2145; Leica Microsystems, Wetzlar, Germany). The 671 sections were stained with Azan-novum according to Geidies using standard protocols (Welsch and 672 Mulisch, 2010). 673 674 Immunohistochemistry 675 676 The brains of fixed animals (4 % FA) were dissected in PBS 0.1 M, pH 7.4, embedded in low- 677 gelling agarose (Cat. A9414; Sigma-Aldrich Chemie GmbH, Munich, Germany) and sectioned (100 µm) 678 with a vibratome (Hyrax V50; Carl Zeiss, Oberkochen, Germany). The sections were preincubated for 679 1.5 h in PBT (PBS + 0.3 % Triton X-100 + 1 % bovine serum albumine) to improve antibody 680 penetration. Two sets of combinations of markers were used: 1. anti-synapsin+ anti-allatostatin+ 681 nuclear marker; 2. anti-synapsin + anti-serotonin+ nuclear marker. The sections were first incubated 682 overnight in the primary antisera at room temperature. The antisera used were: monoclonal anti- 683 SYNORF1 synapsin antibody (DSHB, 3C11; from mouse); polyclonal anti-A-allatostatin antiserum (A- 684 type Dip-allatostatin I; Jena Bioscience, abd-062; from rabbit); polyclonal anti-Serotonin (5-HT, 685 Immunostar, Cat. No 20080, from rabbit, igG). After incubation, the sections were washed in several 686 changes of PBT for 1 h and afterwards incubated in the secondary antibodies (anti IgGs) conjugated 687 to Alexa Fluor 488 (Alexa Fluor 488 goat anti-rabbit IgG Antibody, invitrogen, Molecular Probes) and 688 Cy3 (Cy3-conjugated AffiniPure Goat Anti-Mouse IgG Antibody, Jackson ImmunoResearch 689 Laboratories Inc.) overnight at room temperature. Additionally, HOECHST 33258 (Cat. 14530; Sigma- 690 Aldrich Chemie GmbH, Munich, Germany) was used as a nuclear marker to show the cell clusters. 691 The sections were finally washed in several changes of PBT for 2 h and mounted in Mowiol 4-88 (Cat. 692 0713.2; Carl Roth, Karlsruhe, Germany). 693 694 Antibody specificity 695 696  Synapsin 697 The monoclonal anti-SYNORF1 synapsin antibody (DSHB Hybridoma Product 3C11; anti 698 SYNORF1 as deposited to the DSHB by E. Buchner) was raised against a Drosophila melanogaster 699 GST-synapsin fusion protein and recognizes at least four synapsin isoforms (70, 74, 80 and 143 kDa) 700 in western blots of D. melanogaster head homogenates (Klagges et al., 1996). Sullivan and co- 701 workers (2007) showed a single band at ca. 75 kDa in a western blot analysis of crayfish 702 homogenates. Harzsch and Hansson (2008) conducted a western blot analysis comparing brain tissue 703 of D. melanogaster and the hermit crab Coenobita clypeatus (Anomura, Coenobitidae). The SYNORF1 704 serum provided identical results for both species and it stained one strong band between 80 and 90 705 kDa and a second weaker band slightly above 148 kDa, suggesting that the epitope that SYNORF1 706 recognizes is strongly conserved between D. melanogaster and C. clypeatus (see Harzsch and 707 Hansson, 2008). Similar to the fruit fly, the antibody consistently labels brain structures in other 708 major subgroups of the malacostracan crustaceans (e.g., Beltz et al., 2003; Harzsch et al., 2002, 1999, 709 1998; Krieger et al., 2012) in a pattern that is consistent with the assumption that this antibody labels 710 synaptic neuropils in crustaceans. The antibody also labels neuromuscular synapses in Crustacea and 711 Drosophila (Harzsch et al., 2002). 712 713  Allatostatin 714 The A-type allatostatins represent a large family of neuropeptides that were first identified 715 from the cockroach Diploptera punctata; they additionally share the C-terminal motif -YXFGLamide 716 (Christie et al., 2010; Nässel and Homberg, 2006; Stay et al., 1995; Stay and Tobe, 2007). In the shore 717 crab Carcinus maenas (Brachyura), almost 20 native A-type allatostatin-like peptides were identified 718 from extracts of the thoracic ganglia (Duve et al., 1997). Shortly afterwards, various other A-type 719 allatostatin-like peptides were isolated from the Eastern Crayfish Orconectes limosus (Astacida; 720 Dircksen et al., 1999). Meanwhile, A-type allatostatin peptides have been discovered in a wide range 721 of malacostracan crustaceans, including Brachyura (e.g. Huybrechts et al., 2003), Astacida (e.g. Cape 722 et al., 2008), the prawns Penaeus monodon (Duve et al., 2002), Macrobrachium rosenbergii (Yin et 723 al., 2006) and also in the shrimp Penaeus vannamei (Ma et al., 2010; Meth et al., 2017). Christie 724 (2016) identified a total of 29 peptides with the C-terminal motif, -YXFGLamide, in the latest analysis 725 on the peptidome of the shore crab. The polyclonal rabbit allatostatin antiserum used in the present 726 study was raised against the Diploptera punctata A-type Dip-allatostatin I,APSGAQRLYGFGLamide, 727 coupled to bovine thyroglobulin using glutaraldehyde (Vitzthum et al., 1996). It has previously been 728 used to localize A-type allatostatin-like peptides in crustacean and insect nervous systems (e.g., 729 Kreissl et al., 2010; Polanska et al., 2012). In the following, the term “allatostatin-like 730 immunoreactivity” is used to indicate that the antibody most likely binds to various related peptides 731 within this peptide family. 732 733  Serotonin 734 The antiserum against serotonin (ImmunoStar Incorporated; Cat. No. 20080, Lot No. 541016) 735 is a polyclonal rabbit antiserum raised against serotonin coupled to bovine serum albumin (BSA) with 736 paraformaldehyde. The antiserum was quality control tested by the manufacturer using standard 737 immunohistochemical methods. According to the manufacturer, staining with the antiserum was 738 completely eliminated by pretreatment of the diluted antibody with 25 μg of serotonin coupled to 739 BSA per ml of the diluted antibody. We repeated this control with the serotonin-BSA conjugate that 740 was used for generation of the antiserum as provided by ImmunoStar (Cat. No. 20081, Lot No. 741 750256; 50 μg of lyophilized serotonin creatinine sulfate coupled to BSA with paraformaldehyde). 742 Preadsorption of the antibody in working dilution with the serotonin-BSA conjugate at a final 743 conjugate concentration of 10 μg/ml at 4 °C for 24 h completely blocked all immunolabeling. We 744 performed an additional control and preadsorbed the diluted antiserum with 10 mg/ml BSA for 4 h at 745 room temperature. This preadsorption did not affect the staining, thus, providing evidence that the 746 antiserum does not recognize the carrier molecule alone. The manufacturer also examined the cross 747 reactivity of the antiserum. According to the data sheet, with 5 μg, 10 μg, and 25 μg amounts, the 748 following substances did not react with the antiserum diluted to 1:20,000 using the horse radish 749 peroxidase (HRP) labeling method: 5-hydroxytryptophan, 5-hydroxyindole-3-acetic acid, and 750 dopamine. 751 752 Imaging 753 754 The brain tissues processed for immunofluorescence were viewed with a Leica TCS SP5II 755 confocal laser-scanning microscope equipped with DPSS, Diode- and Argon-lasers and operated by 756 the Leica “Application Suite Advanced Fluorescence” software package (LASAF) (Leica Microsystems, 757 Wetzlar, Germany). Digital images were processed with Adobe Photoshop CS4 or ImageJ. Only global 758 picture enhancement features (brightness and contrast) were used. 759 The head tissues processed for histology were viewed with a Nikon Eclipse 90i upright 760 microscope and bright-field optics (Nikon, Amstelveen, Netherlands). Serial images using a mounted 761 digital camera (Nikon DS-Fi3) were aligned manually with the 3D-reconstruction software Amira 5.6.0 762 (FEI Visualization Science Group, Burlington, VT, USA). 763 For frontal and sagittal sections, dorsal is always towards the top. 764 In the figures, the following color-coded abbreviations were used to identify the markers: 765 SYN, synapsin (magenta); AstA, allatostatin (green); 5HT, serotonin (green); NUC, nuclear counter 766 stain (cyan). Colors were chosen according to Color Universal Design for accessibility to colorblind 767 readers. 768 769 X-ray micro-computed tomography 770 771 Micro-CT scans were performed using an X-ray microscope (Xradia MicroXCT-200; Carl Zeiss 772 Microscopy GmbH, Jena, Germany) that uses a 90-kV/8-W tungsten X-ray source and switchable 773 scintillator-objective lens units as described by Sombke et al. (2015). The heads of fixed animals 774 (Bouin) were contrasted in iodine solution (2 % iodine resublimated (Cat. #X864.1; Carl Roth GmbH, 775 Karlsruhe, Germany) in 99.5 % ethanol), critical point-dried using a fully automatic critical point dryer 776 Leica EM CPD300 (Leica Microsystems, Wetzlar, Germany) and scanned dry (scan medium air). 777 Tomography projections were reconstructed using the reconstruction software XMReconsructor 778 (Carl Zeiss Microscopy GmbH, Jena, Germany), resulting in image stacks (DICOM format) with a pixel 779 size of about 5.8 µm for the 4× objective and 1.9 µm for the 10× objective. 780 781 3D Reconstruction 782 783 The 3D reconstructions of brain and substructures are based on manual segmentation based 784 on image stacks obtained either by the micro-CT scans or by the alignment of serial histological 785 sections, and were performed using the software Amira (FEI Visualization Science Group, Burlington, 786 VT, USA) as described in Sombke et al., 2015. The computed 3D surfaces were slightly smoothed. 787 788 Nomenclature 789 790 The neuroanatomical nomenclature used in this manuscript for neuropils, clusters of cell 791 bodies and tracts is based on Sandeman et al., 1993 and Richter et al. (2010) with some 792 modifications adopted from Krieger et al. (2015) and Loesel et al. (2013). The term “visual neuropils” 793 is used instead of “optic neuropils” as suggested by Krieger et al. (2015). The terms lamina, medulla 794 and lobula are used for the visual neuropils instead of the lamina ganglionaris, medulla externa and 795 medulla interna (Harzsch, 2002). The term “olfactory neuropil” refers to the deutocerebral 796 chemosensory lobe in Loesel et al. (2013) and Krieger et al. (2015). The olfactory globular tract is 797 named the projection neuron tract (PNT) according to Loesel et al. (2013). Cell clusters are referred 798 by their given numbers in parentheses. Because no border was detectable between the cell clusters 799 (9) and (11), they are collectively referred as cluster (9/11) (Krieger et al., 2015), and accordingly are 800 the cell clusters (2) and (3), referred as cluster (2/3). (x) refers likely to the fusion of the cell clusters 801 (12), (13) and (17) according to the nomenclature from Sandeman et al., 1992. 802 803 Calculations 804 805 For the volumes of the HN-TM (hemiellipsoid body and terminal medulla complex in both 806 hemispheres) and the olfactory neuropils relative to the total brain volume, measurements were 807 made from 3D reconstructions of the structures from micro-CT scans using the Amira software. For 808 the brain volume, the brain was delimited where the antennal nerves and the oesophageal 809 connectives separate from the syncerebrum. Additionally, for P. elegans, the volume of the eyestalks 810 which does neither contain any neuropil nor any cell cluster was subtracted to the total brain 811 volume, so that the comparison with R. exoculata (which does not possess eyestalks) was relevant. 812 Same calculations were applied on the Birgus latro data from Krieger et al. 2010. 813 The number of globuli cells (i.e. cell somata in the cell cluster (5)) was determined by 814 estimation of the globuli cell densities in the cell cluster (5), and the total volume of one cell cluster 815 (5). The globuli cell densities were estimated by direct counting of the somata within 0.02 to 0.04 816 mm² paraffin sections of 0.007 mm thickness (1.3x10-4 to 2.8x10-4 mm3), with a density estimated to 817 be approximatively 1.3x106 globui cells per mm3. The total volume of one cell cluster (5) was 818 calculated from 3D-reconstructions with the Amira software. 819 For the volume of the olfactory neuropils and the number of olfactory glomeruli, 820 measurements and estimations were made from sections revealed by synapsin immunoreactivity as 821 described in Beltz et al. (2003). 822 Acknowledgements

823 824 The authors thank the chief scientist of the Bicose 2018 cruise (M-A. Cambon), as well as the 825 captain and crew of the research vessel “Pourquoi pas?” and the HOV Nautile team for sampling the 826 hydrothermal shrimp. We thank B. Shillito and L. Amand for the use of the PERISCOP device, the 827 Ifremer for providing pictures of R. exoculata swarms, E. Becker for the histological sections, P. O. M. 828 Steinhoff for critical point drying the samples, M. Charmantier for feeback on the organ of Bellonci, 829 and C. Wirkner and S. Scholz for their advice on the myoarterial formation. 830 This study was financed by an incoming international stipend from the University of 831 Greifswald, the German Science Foundation; Grant number: DFG INST 292/119-1 FUGG, DFG INST 832 292/120-1 FUGG, and a fellowship grant from InterRidge. 833 834

835 Competing interests

836 837 The authors declare there are no competing interests. 838 Figure legends

839 840 Figure 1. The Alvinocaridid vent shrimp Rimicaris exoculata.

841 A,B. Swarms of thousands of R. exoculata individuals are crowded along the walls of black smoker 842 hydrothermal vents at the TAG vent site (3600 m depth), Mid-Atlantic Ridge 843 (©IFREMER/Nautile6000, BICOSE 2018 cruise). C. Dorsal view of the cephalothorax of R. exoculata, 844 showing voluminous gill chambers covered by the branchiostegites, dorsal eyes (i.e. ocular plate) 845 with two elongated retinae fused in the anterior region, and sensory appendages (antennae 1 and 2). 846 Scale bar = 5 mm. D. Black-white inverted image from an X-ray micro-CT scan showing a dorsal 847 overview of the R. exoculata cephalothorax, with 3D reconstruction of the brain and associated 848 nerves. 849 850 Figure 2. Comparative brain overview in Caridean vent and shallow-water species.

851 Lateral sketches of the brains of the vent shrimp Rimicaris exoculata (A) and the closely-related 852 shallow-water shrimp Palaemon elegans (B), showing the brain position within the cephalothorax, 853 the position of the main nerves and the subdivision of the brain into three neuromeres called proto-, 854 deuto- and tritocerebrum, plus the visual neuropils. In contrast to P. elegans, R. exoculata does not 855 possess eyestalks and the visual neuropils are fused to the median brain, in a dorsoposterior position 856 behind the lateral protocerebrum.

857 858 Figure 3. Overall organization of the brain of R. exoculata.

859 3D reconstruction (A) and schematic representations (B,C) of the brain and neuropils of R. exoculata 860 viewed from a dorsal, slightly anterior direction. The open white arrow points towards anterior of the 861 body axis. In C, the clusters of cell somata associated with the neuropils are shown. The 3D 862 reconstruction is based on an image stack obtained by serial sectioning of paraffin-embedded 863 material. 864 865 Figure 4. Additional views of the brain morphology in R. exoculata.

866 A-D. 3D reconstruction of the brain of R. exoculata in anterior-left (A,B) and left (C, D) views, based 867 on an image stack obtained by serial sectioning of paraffin-embedded material. B and D include the 868 cell clusters. The brain orientation is sketched in the bottom right corners. E,F. Lateral sections of the 869 brain of R. exoculata from micro-CT scans (black-white inverted images). The section’s positions are 870 depicted in the bottom left corners. White asterisk in E indicates the entrance of axons from the cell 871 cluster (5) into the hemiellipsoid body. Black asterisk in F indicates the tract connecting the anterior 872 region of the terminal medulla to the hemiellipsoid body. The open white arrows point towards 873 anterior of the body axis. Scale bars = 100 µm. 874 875 Figure 5. Lateral protocerebrum: the visual neuropils.

876 A-C. Frontal histological sections in the posterior region of the brain, from anterior to posterior, 877 showing the visual neuropils, associated cell clusters, and part of the vascular system. The white 878 arrow head in B shows the fiber tract connecting the medulla to the terminal medulla. D. Sagittal 879 histological section of the brain, showing the eye nerve fibers projecting from the anterodorsal retina 880 to the lamina. White asterisks indicate the entrance of axons from the cell cluster (5) into the 881 hemiellipsoid body intermediate layer. The open white arrow points towards anterior of the body 882 axis. The section’s positions are sketched in the bottom left corners. Scale bars = 100 µm.

883 884 Figure 6. Lateral protocerebrum: the hemiellipsoid body and the terminal medulla

885 A. 3D reconstruction of the lateral protocerebrum (right hemisphere), viewed from an anterior-left 886 perspective, based on an image stack obtained by X-ray micro-CT scan. A conspicuous arcuate tract 887 connects the anterior region of the terminal medulla to the cap region of the hemiellipsoid body (see 888 also in C). B. Schematic representation of the lateral protocerebrum (right hemisphere), viewed from 889 the left. Dotted lines indicate the section’s position in C-F. C-F. Frontal histological sections of the 890 lateral protocerebrum, from anterior to posterior. The hemiellipsoid body and the terminal medulla 891 receive axons from the cell somata in the cell cluster (5) (white asterisks, C-E) and (4) (white 892 arrowhead, D). An arcuate tract connects the terminal medulla to the cap region of the hemiellipsoid 893 body in the anterior region (black asterisk, C). The terminal medulla also connects to the 894 hemiellipsoid body in the middle region, via arborizing fibers (black arrowheads, E). The projection 895 neuron tract enters the hemiellipsoid body in the posterior region (F). G-J’. Horizontal sections of the 896 lateral protocerebrum, from dorsal to ventral, double or triple-labeled for synapsin immunoreactivity 897 (SYN, magenta), allatostatin-like immunoreactivity (AST) or serotonin immunoreactivity (5HT) (both 898 showed in green), and a nuclear marker (NUC, cyan). The inset (H’) shows an enlargement of the 899 hemiellipsoid body neuropil cap region, with microglomeruli. The section’s positions are sketched in 900 the bottom right corners. Black and white open arrows point towards anterior of the body axis. Scale 901 bars = 100 µm (except in H’, scale bar = 50 µm). 902 903 Figure 7. Median protocerebrum. 904 A-C. Black-white inverted images of horizontal sections of the median protocerebrum labeled for 905 synapsin immunoreactivity (A), allatostatin-like immunoreactivity (B) or serotonin immunoreactivity 906 (C). The section’s positions are sketched in the bottom right corners. Black arrows point towards 907 anterior of the body axis. Scale bars = 100 µm.

908 909 Figure 8. Deutocerebrum and tritocerebrum.

910 A,B. Overview of the deutocerebrum and tritocerebrum (frontal histological sections). A is anterior to 911 B. C,D. Sagittal histological sections of the olfactory neuropils. E,F. Horizontal sections of the 912 olfactory neuropil, triple-labeled for synapsin immunoreactivity (SYN, magenta), allatostatin-like 913 immunoreactivity (AST, E,E’) or serotonin immunoreactivity (5HT, F) (green), and a nuclear marker 914 (NUC, cyan). G. Horizontal section of the transversely stratified (white asterisk) lateral antenna 1 915 neuropil, double-labeled for synapsin immunoreactivity (SYN, magenta) and allatostatin-like 916 immunoreactivity (AST, green). H. Sagittal histological section of the tritocerebrum, and part of the 917 deutocerebrum and median protocerebrum. I. Horizontal section of the transversely stratified 918 antenna 2 neuropil, double-labeled for synapsin immunoreactivity (SYN, magenta) and allatostatin- 919 like immunoreactivity (AST, green). The section’s positions are sketched in the bottom corners. Black 920 and white open arrows point towards anterior of the body axis. Scale bars = 100 µm.

921 922 Figure 9. The organ of Bellonci.

923 A-B’. Frontal histological sections of the anterior region of the brain, showing conspicuous onion 924 body-structures from which a nerve tract emanates (A,B), and which are seemingly closely associated 925 to the cerebral vascular system (B’). The section’s positions are sketched in the bottom left corners. 926 C. Anterodorsolateral overview of the cephalothorax from micro-CT scan. Asterisks indicate the 927 position where the organ of Bellonci nerve connects to the cuticle beneath the anterior region of the 928 ocular plate. C’ shows a 3D reconstruction of the brain and the organ of Bellonci nerve in this region. 929 White arrow points towards anterior of the body axis. Scale bars = 100 µm. 930 931 Figure 10. The myoarterial formation and cerebral vascular system.

932 A. Frontal histological section of the myoarterial formation located between the bilateral retina and 933 above the visual neuropils. The section’s position is sketched in the bottom left corner. Scale bar = 934 100 µm. B,C. 3D reconstruction of the myoarterial formation (orange), part of the cerebral vascular 935 system (blue and cyan) and the brain (yellowish), from lateral (B) and anterolateral (C) views, in the 936 cephalothorax. B’ and C’ show higher magnifications of the cerebral vascular system. Dotted arrows 937 indicate structures inside the brain. Open white arrows point towards anterior of the body axis.

938 939 Figure 11. Comparison of the higher integrative centers and olfactory neuropils in several 940 representatives of Malacostraca displayed at the same scale.

941 Sections of the higher integrative centers (i.e. hemiellipsoid body and terminal medulla) (A-G) and 942 horizontal sections of the olfactory neuropil (a-g) labeled with different sets of antibodies (see 943 below), in several malacostracan species: Nebalia herbstii (Aa, Leptostraca, from Kenning et al., 944 2013), Penaeus vannamei (Bb, Dendrobranchiata, from Meth et al., 2017), Saduria entomon (Cc, 945 Isopoda, from Kenning and Harzsch, 2013), Palaemon elegans (Dd) and Rimicaris exoculata (Ee) 946 (Caridea, this study), Carcinus maenas (Ff, Brachyura, from Krieger et al., 2012) and Birgus latro (Gg, 947 Anomala, from Krieger et al., 2010). 948 Markers: a, SYNir; B, SYNir + RFair; b, SYNir + RFair + NUC; Cc, SYNir + 5HTir; A,d,e, SYNir + ASTir; 949 D,E,F-g, SYNir + ASTir + NUC. 950 ASTir, allatostatin-like immunoreactivity (green); NUC, nuclear marker (cyan); RFair, RFamide-like 951 immunoreactivity (green); SYNir, synapsin immunoreactivity (magenta); 5HTir, serotonin 952 immunoreactivity (green). 953 954 Figure 12. Structure of the hemiellipsoid body in several representatives of Malacostraca.

955 The sketches of the hemiellipsoid body structure are displayed in relative size and include 956 representatives of Leptostraca (Nebalia herbstii, Kenning et al., 2013), Stomatopoda 957 (Neogonodactylus oerstedii, Wolff et al., 2017; Gonodactylus bredini, Sullivan and Beltz, 2004), 958 Dendrobranchiata (Penaeus vannamei, Meth et al., 2017; Penaeus duorarum, Sullivan and Beltz, 959 2004), Caridea (Rimicaris exoculata and Palaemon elegans, this study; Palaemonetes pugio, Sullivan 960 and Beltz, 2004), Stenopodidea (Stenopus hispidus, Sullivan and Beltz, 2004 and Krieger et al. 961 unpublished), Achelata (Panulirus argus, Blaustein et al., 1988), Homarida (Homarus americanus, 962 Sullivan and Beltz, 2001), Astacida (Procambarus clarkii, Sullivan and Beltz, 2001), Anomala (Birgus 963 latro, Krieger et al., 2010); Coenobita clypeatus, Wolff et al., 2012; Pagurus bernhardus, Krieger et al., 964 2012), Brachyura (Carcinus maenas, Krieger et al., 2012), Euphausiacea (Meganyctiphanes norvegica, 965 unpublished), Thermosbaenacea (Tethysbaena argentarii, Stegner et al., 2015), Amphipoda 966 (Orchestia cavimana and Niphargus puteanus, Ramm and Scholtz, 2017), Mictacea (Mictocaris 967 halope, Stegner et al., 2015), Spelaeogriphacea (Spelaeogriphus lepidops, Stegner et al., 2015) and 968 Isopoda (Saduria entomon, Kenning and Harzsch, 2013; Idotea emarginata, Stemme et al., 2014). 969 Sketches were made from sections stained using antibody raised against synapsin, except N. 970 puteanus (antibody raised against tubulin), M. norvegica, P. argus and O. cavimana (histological 971 sections), and N. herbstii (optical section). The symbol “?” indicates that the presence of a 972 hemiellipsoid body is uncertain. The phylogram showing phylogenetic relationships of malacostracan 973 crustaceans is modified from Harzsch and Krieger 2018 (therein modified from Sandeman et al., 974 2014, as compiled after Richter and Scholtz, 2001; Scholtz and Richter, 1995; Wirkner and Richter, 975 2010). 976 977 References 978 979 Ache BW. 2002. Crustaceans as animal models for olfactory research. In: Wiese K (Ed.), Crustacean 980 Experimental Systems in Neurobiology. Springer, Berlin, Heidelberg, pp. 189–199. 981 doi:10.1007/978-3-642-56092-7_11 982 Bates AE, Lee RW, Tunnicliffe V, Lamare MD. 2010. Deep-sea hydrothermal vent animals seek cool 983 fluids in a highly variable thermal environment. Nature Communications 1:14. 984 doi:10.1038/ncomms1014 985 Beltz BS, Kordas K, Lee MM, Long JB, Benton JL, Sandeman DC. 2003. Ecological, evolutionary, and 986 functional correlates of sensilla number and glomerular density in the olfactory system of 987 decapod crustaceans. Journal of Comparative Neurology 455:260–269. doi:10.1002/cne.10474 988 Bengochea M, Berón de Astrada M, Tomsic D, Sztarker J. 2017. A crustacean lobula plate: 989 Morphology, connections, and retinotopic organization. Journal of Comparative Neurology 990 526:109–119. doi:10.1002/cne.24322 991 Blaustein DN, Derby CD, Simmons RB, Beall AC. 1988. Structure of the brain and medulla terminalis 992 of the spiny lobster Panulirus argus and the crayfish Procambarus clarkii, with an emphasis on 993 olfactory centers. Journal of Crustacean Biology 8:493–519. doi:10.1163/193724088X00341 994 Breithaupt T, Thiel M, editors. 2011. Chemical Communication in Crustaceans. New York: Springer- 995 Verlag. 996 Brenneis G, Richter S. 2010. Architecture of the nervous system in Mystacocarida (Arthropoda, 997 Crustacea)—an immunohistochemical study and 3D reconstruction. Journal of morphology 998 271:169–189. 999 Brown S, Wolff G. 2012. Fine structural organization of the hemiellipsoid body of the land hermit 1000 crab, Coenobita clypeatus. Journal of Comparative Neurology 520:2847–2863. 1001 doi:10.1002/cne.23058 1002 Cape SS, Rehm KJ, Ma M, Marder E, Li L. 2008. Mass spectral comparison of the neuropeptide 1003 complement of the stomatogastric ganglion and brain in the adult and embryonic lobster, 1004 Homarus americanus. Journal of neurochemistry 105:690–702. 1005 Casanova B, Brunet M, Segonzac M. 1993. Bacterian epibiosis impact on the morphology of shrimps 1006 associated with hydrothermalism in the Mid-Atlantic. Cahiers de Biologie Marine 34:573–588. 1007 Cathalot C, Rouxel O, et al. 2018. Rapport de mission BICOSE2 - partie géochimie / Scientifique et 1008 Instrumentation. 1009 Chaigneau J. 1994. Les organes des sens autres que ceux de la vision. Traité de Zoologie 7:345–410. 1010 Chamberlain SC. 2000. Vision in hydrothermal vent shrimp. Philosophical Transactions of the Royal 1011 Society of London B: Biological Sciences 355:1151–1154. doi:10.1098/rstb.2000.0657 1012 Charlou JL, Donval JP, Douville E, Jean-Baptiste P, Radford-Knoery J, Fouquet Y, Dapoigny A, 1013 Stievenard M. 2000. Compared geochemical signatures and the evolution of Menez Gwen 1014 (37°50′N) and Lucky Strike (37°17′N) hydrothermal fluids, south of the Azores Triple Junction on 1015 the Mid-Atlantic Ridge. Chemical Geology 171:49–75. doi:10.1016/S0009-2541(00)00244-8 1016 Charlou JL, Donval JP, Fouquet Y, Jean-Baptiste P, Holm N. 2002. Geochemistry of high H2 and CH4 1017 vent fluids issuing from ultramafic rocks at the Rainbow hydrothermal field (36°14′N, MAR). 1018 Chemical Geology 191:345–359. doi:10.1016/S0009-2541(02)00134-1 1019 Charlou JL, Donval JP, Konn C, Ondréas H, Fouquet Y, Jean‐Baptiste P, Fourré E. 2010. High 1020 production and fluxes of H2 and CH4 and evidence of abiotic hydrocarbon synthesis by 1021 serpentinization in ultramafic-hosted hydrothermal systems on the Mid-Atlantic Ridge. 1022 Geophysical Monograph Series 188:265-296. doi:10.1029/2008GM000752 1023 Charmantier-Daures M, Segonzac M. 1998. Organ of Bellonci and sinus gland in three decapods from 1024 Atlantic hydrothermal vents: Rimicaris Exoculata, Chorocaris Chacei, and Segonzacia 1025 Mesatlantica. Journal of Crustacean Biology 18:213–223. doi:10.2307/1549315 1026 Chausson F, Sanglier S, Leize E, Hagège A, Bridges CR, Sarradin P-M, Shillito B, Lallier FH, Zal F. 2004. 1027 Respiratory adaptations to the deep-sea hydrothermal vent environment: the case of Segonzacia 1028 mesatlantica, a crab from the Mid-Atlantic Ridge. Micron, XIIIth International Conference on 1029 Invertebrate Dioxygen Binding Proteins 35:31–41. doi:10.1016/j.micron.2003.10.010 1030 Chaves da Silva PG, Benton JL, Sandeman DC, Beltz BS. 2012. Adult Neurogenesis in the Crayfish 1031 Brain: The Hematopoietic Anterior Proliferation Center Has Direct Access to the Brain and Stem 1032 Cell Niche. Stem Cells and Development 22:1027–1041. doi:10.1089/scd.2012.0583 1033 Childress JJ, Fisher CR. 1992. The biology of hydrothermal vent animals: physiology, biochemistry, 1034 and autotrophic symbioses. Oceanography and Marine Biology: An Annual Review 30:337–441. 1035 Christie AE. 2016. Expansion of the neuropeptidome of the globally invasive marine crab Carcinus 1036 maenas. General and comparative endocrinology 235:150–169. 1037 Christie AE, Stemmler EA, Dickinson PS. 2010. Crustacean neuropeptides. Cellular and Molecular Life 1038 Sciences 67:4135–4169. 1039 Corbari L, Zbinden M, Cambon-Bonavita M-A, Gaill F, Compère P. 2008. Bacterial symbionts and 1040 mineral deposits in the branchial chamber of the hydrothermal vent shrimp Rimicaris exoculata: 1041 relationship to moult cycle. Aquatic Biology 1:225–238. 1042 Cottin D, Shillito B, Chertemps T, Thatje S, Léger N, Ravaux J. 2010. Comparison of heat-shock 1043 responses between the hydrothermal vent shrimp Rimicaris exoculata and the related coastal 1044 shrimp Palaemonetes varians. Journal of Experimental Marine Biology and Ecology 393:9–16. 1045 doi:10.1016/j.jembe.2010.06.008 1046 Cronin TW, Porter ML. 2008. Exceptional variation on a common theme: the evolution of crustacean 1047 compound eyes. Evolution: Education and Outreach 1:463. 1048 Derby CD, Thiel M. 2014. Nervous Systems and Control of Behavior. Oxford University Press. 1049 Derby CD, Weissburg MJ. 2014. The chemical senses and chemosensory ecology of crustaceans. In: 1050 Derby CD, Thiel M (Eds.), Nervous Systems and Control of Behavior: Vol. 3. Natural History of 1051 Crustacea. Oxford University Press, New York, pp. 263–292. 1052 Derby CD, Fortier JK, Harrison PJH, Cate HS. 2003. The peripheral and central antennular pathway of 1053 the Caribbean stomatopod crustacean Neogonodactylus oerstedii. Arthropod Structure & 1054 Development 32:175–188. doi:10.1016/S1467-8039(03)00048-3 1055 Desbruyères D, Almeida A, Biscoito M, Comtet T, Khripounoff A, Bris NL, Sarradin PM, Segonzac M. 1056 2000. A review of the distribution of hydrothermal vent communities along the northern Mid- 1057 Atlantic Ridge: dispersal vs. environmental controls. Hydrobiologia 440:201-216. doi:10.1007/978- 1058 94-017-1982-7_19 1059 Desbruyères D, Biscoito M, Caprais J-C, Colaço A, Comtet T, Crassous P, Fouquet Y, Khripounoff A, Le 1060 Bris N, Olu K, Riso R, Sarradin P-M, Segonzac M, Vangriesheim A. 2001. Variations in deep-sea 1061 hydrothermal vent communities on the Mid-Atlantic Ridge near the Azores plateau. Deep Sea 1062 Research Part I: Oceanographic Research Papers 48:1325–1346. doi:10.1016/S0967- 1063 0637(00)00083-2 1064 Dircksen H, Skiebe P, Abel B, Agricola H, Buchner K, Muren JE, Nässel DR. 1999. Structure, 1065 distribution, and biological activity of novel members of the allatostatin family in the crayfish 1066 Orconectes limosus. Peptides 20:695–712. 1067 Duffy JE, Thiel M, editors. 2007. Evolutionary Ecology of Social and Sexual Systems: Crustaceans as 1068 Model Organisms. Oxford, New York: Oxford University Press. 1069 Duve H, Johnsen AH, Maestro J-L, Scott AG, Jaros PP, Thorpe A. 1997. Isolation and identification of 1070 multiple neuropeptides of the allatostatin superfamily in the shore crab Carcinus maenas. 1071 European Journal of Biochemistry 250:727–734. 1072 Duve H, Johnsen AH, Scott AG, Thorpe A. 2002. Allatostatins of the tiger prawn, Penaeus monodon 1073 (Crustacea: Penaeidea). Peptides 23:1039–1051. 1074 Elofsson R, Hessler RR. 1990. Central nervous system of Hutchinsoniella macracantha 1075 (Cephalocarida). Journal of Crustacean Biology 10:423–439. 1076 Fanenbruck M, Harzsch S. 2005. A brain atlas of Godzilliognomus frondosus Yager, 1989 (Remipedia, 1077 Godzilliidae) and comparison with the brain of Speleonectes tulumensis Yager, 1987 (Remipedia, 1078 Speleonectidae): implications for arthropod relationships. Arthropod Structure & Development, 1079 Arthropod Brain Morphology and Evolution 34:343–378. doi:10.1016/j.asd.2005.01.007 1080 Fanenbruck M, Harzsch S, Wägele JW. 2004. The brain of the Remipedia (Crustacea) and an 1081 alternative hypothesis on their phylogenetic relationships. Proceedings of the National Academy 1082 of Sciences 101:3868–3873. doi:10.1073/pnas.0306212101 1083 Fisher CR, Takai K, Le Bris N. 2007. Hydrothermal Vent Ecosystems. Oceanography 20:14–23. 1084 Fontaine MT, Passelecq-Gerin E, Bauchau AG. 1982. Structures chemoreceptrices des antennules du 1085 crabe Carcinus Maenas (L.) (Decapoda Brachyura). Crustaceana 43:271–283. 1086 doi:10.1163/156854082X00218 1087 Garm A, Hallberg E, Høeg JT. 2003. Role of maxilla 2 and its setae during feeding in the shrimp 1088 Palaemon adspersus (Crustacea: Decapoda). The Biological Bulletin 204:126–137. 1089 doi:10.2307/1543548 1090 Garm A, Shabani S, Høeg JT, Derby CD. 2005. Chemosensory neurons in the mouthparts of the spiny 1091 lobsters Panulirus argus and Panulirus interruptus (Crustacea: Decapoda). Journal of Experimental 1092 Marine Biology and Ecology 314:175–186. doi:10.1016/j.jembe.2004.08.016 1093 Garm A, Watling L. 2013. The crustacean integument: setae, setules, and other ornamentation. 1094 Functional Morphology and Diversity 1:167–198. 1095 Gaten E, Herring PJ, Shelton PA. 1998. The development and evolution of the eyes of vent shrimps 1096 (Decapoda: Bresiliidae). Cahiers de Biologie Marine 39:287–290. 1097 Gebruk AV, Galkin SV, Vereshchaka AL, Moskalev LI, Southward AJ. 1997. Ecology and biogeography 1098 of the hydrothermal vent fauna of the Mid-Atlantic Ridge. In: Blaxter JHS, Southward AJ, Gebruk 1099 AV, Southward EC, Tyler PA (Eds.), Advances in Marine Biology, The Biogeography of the Oceans. 1100 Academic Press. pp. 93–144. doi:10.1016/S0065-2881(08)60016-4 1101 Grünert U, Ache BW. 1988. Ultrastructure of the aesthetasc (olfactory) sensilla of the spiny lobster, 1102 Panulirus argus. Cell Tissue Res 251:95–103. doi:10.1007/BF00215452 1103 Guenther CM, Atema J. 1998. Distribution of setae on the Homarus americanus lateral antennular 1104 flagellum. The Biological Bulletin 195:182–183. doi:10.2307/1542827 1105 Hallberg E, Skog M. 2011. Chemosensory sensilla in crustaceans. In: Breithaupt T, Thiel M (Eds.), 1106 Chemical Communication in Crustaceans. Springer, New York, pp. 103–121. 1107 Harzsch S. 2002. The phylogenetic significance of crustacean optic neuropils and chiasmata: a re- 1108 examination. Journal of Comparative Neurology 453:10–21. 1109 Harzsch S, Anger K, Dawirs RR. 2002. Immunocytochemical detection of acetylated alpha-tubulin and 1110 Drosophila synapsin in the embryonic crustacean nervous system. International Journal of 1111 Developmental Biology 41:477–484. 1112 Harzsch S, Benton J, Dawirs RR, Beltz B. 1999. A new look at embryonic development of the visual 1113 system in decapod crustaceans: neuropil formation, neurogenesis, and apoptotic cell death. 1114 Journal of neurobiology 39:294. 1115 Harzsch S, Hansson BS. 2008. Brain architecture in the terrestrial hermit crab Coenobita clypeatus 1116 (Anomura, Coenobitidae), a crustacean with a good aerial sense of smell. BMC Neuroscience 9:58. 1117 doi:10.1186/1471-2202-9-58 1118 Harzsch S, Krieger J. 2018. Crustacean olfactory systems: A comparative review and a crustacean 1119 perspective on olfaction in insects. Progress in Neurobiology 161:23–60. 1120 doi:10.1016/j.pneurobio.2017.11.005 1121 Harzsch S, Miller J, Benton J, Dawirs RR, Beltz B. 1998. Neurogenesis in the thoracic neuromeres of 1122 two crustaceans with different types of metamorphic development. Journal of Experimental 1123 Biology 201:2465–2479. 1124 Harzsch S, Rieger V, Krieger J, Seefluth F, Strausfeld NJ, Hansson BS. 2011. Transition from marine to 1125 terrestrial ecologies: changes in olfactory and tritocerebral neuropils in land-living isopods. 1126 Arthropod structure & development 40:244–257. 1127 Herring PJ, Gaten E, Shelton PMJ. 1999. Are vent shrimps blinded by science? Nature 398:116. 1128 doi:10.1038/18142 1129 Hourdez S, Lallier FH. 2006. Adaptations to hypoxia in hydrothermal-vent and cold-seep 1130 invertebrates. In: Life in Extreme Environments. Springer. pp. 297–313. 1131 Hügler M, Petersen JM, Dubilier N, Imhoff JF, Sievert SM. 2011. Pathways of carbon and energy 1132 metabolism of the epibiotic community associated with the deep-sea hydrothermal vent shrimp 1133 Rimicaris exoculata. PLoS One 6:e16018. 1134 Huybrechts J, Nusbaum MP, Bosch LV, Baggerman G, De Loof A, Schoofs L. 2003. Neuropeptidomic 1135 analysis of the brain and thoracic ganglion from the Jonah crab, Cancer borealis. Biochemical and 1136 biophysical research communications 308:535–544. 1137 Jannasch HW, Mottl MJ. 1985. Geomicrobiology of deep-sea hydrothermal vents. Science 229:717– 1138 725. doi:10.1126/science.229.4715.717 1139 Jinks RN, Battelle B-A, Herzog ED, Kass L, Renninger GH, Chamberlain SC. 1998. Sensory adaptations 1140 in hydrothermal vent shrimps from the Mid-Atlantic Ridge. Cahiers de Biologie Marine 39:309– 1141 312. 1142 Johnson KS, Beehler CL, Sakamoto-Arnold CM, Childress JJ. 1986. In situ measurements of chemical 1143 distributions in a deep-sea hydrothermal vent field. Science 231:1139–1141. 1144 doi:10.1126/science.231.4742.1139 1145 Johnson KS, Childress JJ, Beehler CL. 1988. Short-term temperature variability in the Rose Garden 1146 hydrothermal vent field: an unstable deep-sea environment. Deep Sea Research Part A 1147 Oceanographic Research Papers 35:1711–1721. doi:10.1016/0198-0149(88)90045-3 1148 Johnson ML, Shelton PMJ, Herring PJ, Gardner S. 1995. Spectral responses from the dorsal organ of a 1149 juvenile Rimicaris exoculata from the TAG hydrothermal vent site. BRIDGE Newsletter 8:38–42. 1150 Jr DFM, Alones VE. 1997. Response properties of higher level neurons in the central olfactory 1151 pathway of the crayfish. Journal of Comparative Physiology A 181:205–216. 1152 doi:10.1007/s003590050107 1153 Kenning M, Harzsch S. 2013. Brain anatomy of the marine isopod Saduria entomon Linnaeus, 1758 1154 (Valvifera, Isopoda) with special emphasis on the olfactory pathway. Frontiers in neuroanatomy 1155 7:32. 1156 Kenning M, Müller C, Wirkner CS, Harzsch S. 2013. The Malacostraca (Crustacea) from a 1157 neurophylogenetic perspective: New insights from brain architecture in Nebalia herbstii Leach, 1158 1814 (Leptostraca, Phyllocarida). Zoologischer Anzeiger - A Journal of Comparative Zoology 1159 252:319–336. doi:10.1016/j.jcz.2012.09.003 1160 Klagges BR, Heimbeck G, Godenschwege TA, Hofbauer A, Pflugfelder GO, Reifegerste R, Reisch D, 1161 Schaupp M, Buchner S, Buchner E. 1996. Invertebrate synapsins: a single gene codes for several 1162 isoforms in Drosophila. Journal of Neuroscience 16:3154–3165. 1163 Klaus S, Mendoza JC, Liew JH, Plath M, Meier R, Yeo DC. 2013. Rapid evolution of troglomorphic 1164 characters suggests selection rather than neutral mutation as a driver of eye reduction in cave 1165 crabs. Biology letters 9:20121098. 1166 Kreissl S, Strasser C, Galizia CG. 2010. Allatostatin immunoreactivity in the honeybee brain. Journal of 1167 Comparative Neurology 518:1391–1417. 1168 Krieger J, Braun P, Rivera NT, Schubart CD, Müller CHG, Harzsch S. 2015. Comparative analyses of 1169 olfactory systems in terrestrial crabs (Brachyura): evidence for aerial olfaction? PeerJ 3:e1433. 1170 doi:10.7717/peerj.1433 1171 Krieger J, Grandy R, Drew MM, Erland S, Stensmyr MC, Harzsch S, Hansson BS. 2012a. Giant Robber 1172 Crabs Monitored from Space: GPS-Based Telemetric Studies on Christmas Island (Indian Ocean). 1173 PLOS ONE 7:e49809. doi:10.1371/journal.pone.0049809 1174 Krieger J, Sandeman RE, Sandeman DC, Hansson BS, Harzsch S. 2010. Brain architecture of the largest 1175 living land arthropod, the Giant Robber Crab Birgus latro (Crustacea, Anomura, Coenobitidae): 1176 evidence for a prominent central olfactory pathway? Frontiers in Zoology 7:25. doi:10.1186/1742- 1177 9994-7-25 1178 Krieger J, Sombke A, Seefluth F, Kenning M, Hansson BS, Harzsch S. 2012b. Comparative brain 1179 architecture of the European shore crab Carcinus maenas (Brachyura) and the common hermit 1180 crab Pagurus bernhardus (Anomura) with notes on other marine hermit crabs. Cell and Tissue 1181 Research 348:47–69. doi:10.1007/s00441-012-1353-4 1182 Lallier FH, Truchot J-P. 1997. Hemocyanin oxygen-binding properties of a deep-sea hydrothermal 1183 vent shrimp: Evidence for a novel cofactor. Journal of Experimental Zoology 277:357–364. 1184 doi:10.1002/(SICI)1097-010X(19970401)277:5<357::AID-JEZ1>3.0.CO;2-O 1185 Le Bris N, Zbinden M, Gaill F. 2005. Processes controlling the physico-chemical micro-environments 1186 associated with Pompeii worms. Deep Sea Research Part I: Oceanographic Research Papers 1187 52:1071–1083. doi:10.1016/j.dsr.2005.01.003 1188 Loesel R, Wolf H, Kenning M, Harzsch S, Sombke A. 2013. Architectural principles and evolution of 1189 the Arthropod central nervous system. In: Minelli A, Boxhsall G, Fusco G (Eds.), Arthropod Biology 1190 and Evolution. Berlin, Heidelberg, pp. 299–342. doi:10.1007/978-3-642-36160-9_13 1191 Lunina AA, Vereshchaka AL. 2014. Distribution of hydrothermal alvinocaridid shrimps: effect of 1192 geomorphology and specialization to extreme biotopes. PloS one 9:e92802. 1193 Luther III GW, Rozan TF, Taillefert M, Nuzzio DB, Di Meo C, Shank TM, Lutz RA, Cary SC. 2001. 1194 Chemical speciation drives hydrothermal vent ecology. Nature 410:813. 1195 Ma M, Gard AL, Xiang F, Wang J, Davoodian N, Lenz PH, Malecha SR, Christie AE, Li L. 2010. 1196 Combining in silico transcriptome mining and biological mass spectrometry for neuropeptide 1197 discovery in the Pacific white shrimp Litopenaeus vannamei. Peptides 31:27–43. 1198 Machon J, Lucas P, Ravaux J, Zbinden M. 2018. Comparison of chemoreceptive abilities of the 1199 hydrothermal shrimp Mirocaris fortunata and the coastal shrimp Palaemon elegans. Chemical 1200 senses 43(7):489-501. doi:https://doi.org/10.1093/chemse/bjy041 1201 Maza FJ, Sztarker J, Shkedy A, Peszano VN, Locatelli FF, Delorenzi A. 2016. Context-dependent 1202 memory traces in the crab’s mushroom bodies: Functional support for a common origin of high- 1203 order memory centers. PNAS 113:E7957–E7965. doi:10.1073/pnas.1612418113 1204 McGaw IJ. 2005. The decapod crustacean circulatory system: A case that is neither open nor closed. 1205 Microscopy and Microanalysis 11:18–36. doi:10.1017/S1431927605050026 1206 McGaw IJ, Reiber CL. 2002. Cardiovascular system of the blue crab Callinectes sapidus. Journal of 1207 Morphology 251:1–21. doi:10.1002/jmor.1071 1208 McKinzie ME, Benton JL, Beltz BS, Mellon D. 2003. Parasol cells of the hemiellipsoid body in the 1209 crayfish Procambarus clarkii: Dendritic branching patterns and functional implications. Journal of 1210 Comparative Neurology 462:168–179. doi:10.1002/cne.10716 1211 McMahon BR. 2001. Control of cardiovascular function and its evolution in Crustacea. Journal of 1212 Experimental Biology 204:923–932. 1213 Mellon D. 2014. Sensory systems of crustaceans. In: Derby C, Thiel M, Thiel M (Eds.), The Natural 1214 History of the Crustacea 3:49–84. 1215 Mellon D. 2012. Smelling, feeling, tasting and touching: behavioral and neural integration of 1216 antennular chemosensory and mechanosensory inputs in the crayfish. Journal of Experimental 1217 Biology 215:2163–2172. doi:10.1242/jeb.069492 1218 Mellon D. 2003. Dendritic initiation and propagation of spikes and spike bursts in a multimodal 1219 sensory interneuron: The crustacean parasol cell. Journal of Neurophysiology 90:2465–2477. 1220 doi:10.1152/jn.00310.2003 1221 Mellon D. 2000. Convergence of ultimodal sensory input onto higher-level neurons of the crayfish 1222 olfactory pathway. Journal of Neurophysiology 84:3043–3055. doi:10.1152/jn.2000.84.6.3043 1223 Mellon D, Sandeman DC, Sandeman RE. 1992. Characterization of oscillatory olfactory interneurones 1224 in the protocerebrum of the crayfish. Journal of Experimental Biology 167:15–38. 1225 Mellon D, Wheeler CJ. 1999. Coherent oscillations in membrane potential synchronize impulse bursts 1226 in central olfactory neurons of the crayfish. Journal of Neurophysiology 81:1231–1241. 1227 doi:10.1152/jn.1999.81.3.1231 1228 Mellon D, Alones V, Lawrence MD. 1992. Anatomy and fine structure of neurons in the 1229 deutocerebral projection pathway of the crayfish olfactory system. Journal of Comparative 1230 Neurology 321:93–111. doi:10.1002/cne.903210109 1231 Meth R, Wittfoth C, Harzsch S. 2017. Brain architecture of the Pacific White Shrimp Penaeus 1232 vannamei Boone, 1931 (Malacostraca, Dendrobranchiata): correspondence of brain structure and 1233 sensory input? Cell and Tissue Research 369:255–271. doi:10.1007/s00441-017-2607-y 1234 Moran D, Softley R, Warrant EJ. 2015. The energetic cost of vision and the evolution of eyeless 1235 Mexican cavefish. Science advances 1:e1500363. 1236 Nässel DR, Homberg U. 2006. Neuropeptides in interneurons of the insect brain. Cell and Tissue 1237 Research 326:1. 1238 Niven JE, Laughlin SB. 2008. Energy limitation as a selective pressure on the evolution of sensory 1239 systems. Journal of Experimental Biology 211:1792–1804. doi:10.1242/jeb.017574 1240 Nuckley DJ, Jinks RN, Battelle BA, Herzog ED, Kass L, Renninger GH, Chamberlain SC. 1996. Retinal 1241 anatomy of a new species of bresiliid shrimp from a hydrothermal vent field on the Mid-Atlantic 1242 Ridge. The Biological Bulletin 190:98–110. doi:10.2307/1542679 1243 O’Neill PJ, Jinks RN, Herzog ED, Battelle B-A, Kass L, Renninger GH, Chamberlain SC. 1995. The 1244 morphology of the dorsal eye of the hydrothermal vent shrimp, Rimicaris exoculata. Visual 1245 Neuroscience 12:861–875. doi:10.1017/S0952523800009421 1246 Pelli DG, Chamberlain SC. 1989. The visibility of 350 °C black-body radiation by the shrimp Rimicaris 1247 exoculata and man. Nature 337:460–461. doi:10.1038/337460a0 1248 Petersen JM, Ramette A, Lott C, Cambon‐Bonavita M-A, Zbinden M, Dubilier N. 2010. Dual symbiosis 1249 of the vent shrimp Rimicaris exoculata with filamentous gamma- and epsilonproteobacteria at 1250 four Mid-Atlantic Ridge hydrothermal vent fields. Environmental Microbiology 12:2204–2218. 1251 doi:10.1111/j.1462-2920.2009.02129.x 1252 Polanska MA, Tuchina O, Agricola H, Hansson BS, Harzsch S. 2012. Neuropeptide complexity in the 1253 crustacean central olfactory pathway: immunolocalization of A-type allatostatins and RFamide- 1254 like peptides in the brain of a terrestrial hermit crab. Molecular Brain 5:29. doi:10.1186/1756- 1255 6606-5-29 1256 Polanska MA, Yasuda A, Harzsch S. 2007. Immunolocalisation of crustacean-SIFamide in the median 1257 brain and eyestalk neuropils of the marbled crayfish. Cell and Tissue Research 330:331–344. 1258 Ponsard J, Cambon-Bonavita M-A, Zbinden M, Lepoint G, Joassin A, Corbari L, Shillito B, Durand L, 1259 Cueff-Gauchard V, Compère P. 2013. Inorganic carbon fixation by chemosynthetic ectosymbionts 1260 and nutritional transfers to the hydrothermal vent host-shrimp Rimicaris exoculata. The ISME 1261 Journal 7:96–109. doi:10.1038/ismej.2012.87 1262 Pynnönen K. 1985. The structure of long distance (antennular) chemoreceptors in Saduria entomon 1263 (L.), Isopoda, and their role in feeding behaviour. Annales Zoologici Fennici 22:423–432. 1264 Ramm T, Scholtz G. 2017. No sight, no smell? – Brain anatomy of two amphipod crustaceans with 1265 different lifestyles. Arthropod Structure & Development 46:537–551. 1266 doi:10.1016/j.asd.2017.03.003 1267 Rathke MH. 1837. Bertrag zur fauna der krym. Mémoires de l’Académie Impériale des Sciences de St 1268 Petersbourg 3:371–380. 1269 Ravaux J, Gaill F, Bris NL, Sarradin P-M, Jollivet D, Shillito B. 2003. Heat-shock response and 1270 temperature resistance in the deep-sea vent shrimp Rimicaris exoculata. Journal of Experimental 1271 Biology 206:2345–2354. doi:10.1242/jeb.00419 1272 Reiber CL, McMahon BR. 1998. The effects of progressive hypoxia on the crustacean cardiovascular 1273 system: a comparison of the freshwater crayfish, (Procambarus clarkii), and the lobster (Homarus 1274 americanus). Journal of Comparative Physiology B 168:168–176. doi:10.1007/s003600050133 1275 Renninger GH, Kass L, Gleeson RA, Van Dover CL, Battelle B-A, Jinks RN, Herzog ED, Chamberlain SC. 1276 1995. Sulfide as a chemical stimulus for deep-sea hydrothermal vent shrimp. The Biological 1277 Bulletin 189:69–76. doi:10.2307/1542456 1278 Reynolds GT, Lutz RA. 2001. Sources of light in the deep ocean. Reviews of Geophysics 39:123–136. 1279 Richter S, Loesel R, Purschke G, Schmidt-Rhaesa A, Scholtz G, Stach T, Vogt L, Wanninger A, Brenneis 1280 G, Döring C. 2010. Invertebrate neurophylogeny: suggested terms and definitions for a 1281 neuroanatomical glossary. Frontiers in Zoology 7:29. 1282 Richter S, Scholtz G. 2001. Phylogenetic analysis of the Malacostraca (Crustacea). Journal of 1283 Zoological Systematics and Evolutionary Research 39:113–136. doi:10.1046/j.1439- 1284 0469.2001.00164.x 1285 Rona PA, Klinkhammer G, Nelsen TA, Trefry JH, Elderfield H. 1986. Black smokers, massive sulphides 1286 and vent biota at the Mid-Atlantic Ridge. Nature 321:33. 1287 Sandeman D, Callan Harold Garnet. 1967. The vascular circulation in the brain, optic lobes and 1288 thoracic ganglia of the crab Carcinus. Proceedings of the Royal Society of London Series B 1289 Biological Sciences 168:82–90. doi:10.1098/rspb.1967.0052 1290 Sandeman D, Kenning M, Harzsch S. 2014a. Adaptive trends in malacostracan brain form and 1291 function related to behavior. Crustacean Nervous System and Their Control of Behaviour, The 1292 Natural History of the Crustacea 3:11–48. 1293 Sandeman D, Kenning M, Harzsch S. 2014b. Adaptive trends in malacostracan brain form and 1294 function related to behavior. In: Derby C, Thiel M (Eds.), Crustacean Nervous System and Their 1295 Control of Behavior, Vol. 3. Oxford University Press, New York, pp. 11–48. 1296 Sandeman D, Sandeman R, Derby C, Schmidt M. 1992. Morphology of the Brain of Crayfish, Crabs, 1297 and Spiny Lobsters: A Common Nomenclature for Homologous Structures. The Biological Bulletin 1298 183:304–326. doi:10.2307/1542217 1299 Sandeman D, Scholtz G, Sandeman RE. 1993. Brain evolution in decapod crustacea. Journal of 1300 Experimental Zoology 265:112–133. doi:10.1002/jez.1402650204 1301 Sanders NK, Arp AJ, Childress JJ. 1988. Oxygen binding characteristics of the hemocyanins of two 1302 deep-sea hydrothermal vent crustaceans. Respiration Physiology 71:57–67. doi:10.1016/0034- 1303 5687(88)90115-6 1304 Schachtner J, Schmidt M, Homberg U. 2005. Organization and evolutionary trends of primary 1305 olfactory brain centers in Tetraconata (Crustacea+Hexapoda). Arthropod Structure & 1306 Development, Arthropod Brain Morphology and Evolution 34:257–299. 1307 doi:10.1016/j.asd.2005.04.003 1308 Schmidt C, Le Bris N, Gaill F. 2008. Interactions of deep-sea vent invertebrates with their 1309 environment: The case of Rimicaris exoculata. Journal of Shellfish Research 27:79–90. 1310 doi:10.2983/0730-8000(2008)27[79:IODVIW]2.0.CO;2 1311 Schmidt M. 2016. Malacostraca. In: Schmidt-Rhaesa A, Harzsch S, Purschke G (Eds.), Structure & 1312 Evolution of Invertebrate Nervous Systems. Oxford University Press, Oxford, pp. 529–582. 1313 Schmidt M, Ache BW. 1997. Immunocytochemical analysis of glomerular regionalization and 1314 neuronal diversity in the olfactory deutocerebrum of the spiny lobster. Cell and Tissue Research 1315 287:541–563. doi:10.1007/s004410050778 1316 Schmidt M, Gnatzy W. 1984. Are the funnel-canal organs the ‘campaniform sensilla’ of the shore 1317 crab, Carcinus maenas (Decapoda, Crustacea)? Cell and Tissue Research 237:81–93. 1318 doi:10.1007/BF00229202 1319 Schmidt M, Mellon D. 2010. Neuronal processing of chemical information in crustaceans. In: 1320 Breithaupt T, Thiel M (Eds.), Chemical Communication in Crustaceans. Springer, New York, pp. 1321 123–147. doi:10.1007/978-0-387-77101-4_7 1322 Scholtz G, Richter S. 1995. Phylogenetic systematics of the reptantian Decapoda (Crustacea, 1323 Malacostraca). Zoological Journal of the Linnean Society 113:289–328. 1324 Scholz S, Richter S, Wirkner CS. 2018. Constant morphological patterns in the hemolymph vascular 1325 system of crayfish (Crustacea, Decapoda). Arthropod Structure & Development 47:248–267. 1326 doi:10.1016/j.asd.2017.12.005 1327 Segonzac M, De Saint Laurent M, Casanova B. 1993. L’énigme du comportement trophique des 1328 crevettes Alvinocarididae des sites hydrothermaux de la dorsale médio-atlantique. Cahiers De 1329 Biologie Marine 34:535–571. 1330 Shillito B, Hamel G, Duchi C, Cottin D, Sarrazin J, Sarradin P-M, Ravaux J, Gaill F. 2008. Live capture of 1331 megafauna from 2300m depth, using a newly designed Pressurized Recovery Device. Deep Sea 1332 Research Part I: Oceanographic Research Papers 55:881–889. doi:10.1016/j.dsr.2008.03.010 1333 Sinakevitch I, Douglass JK, Scholtz G, Loesel R, Strausfeld NJ. 2003. Conserved and convergent 1334 organization in the optic lobes of insects and isopods, with reference to other crustacean taxa. 1335 Journal of Comparative Neurology 467:150–172. 1336 Sitaraman D, Zars M, LaFerriere H, Chen Y-C, Sable-Smith A, Kitamoto T, Rottinghaus GE, Zars T. 1337 2008. Serotonin is necessary for place memory in Drosophila. Proceedings of the National 1338 Academy of Sciences 105:5579–5584. 1339 Sombke A, Lipke E, Michalik P, Uhl G, Harzsch S. 2015. Potential and limitations of X-Ray micro- 1340 computed tomography in arthropod neuroanatomy: A methodological and comparative survey. 1341 Journal of Comparative Neurology 523:1281–1295. 1342 Stay B, Tobe SS. 2007. The role of allatostatins in juvenile hormone synthesis in insects and 1343 crustaceans. Annual Review of Entomology 52:277–299. 1344 Stay B, Tobe SS, Bendena WG. 1995. Allatostatins: identification, primary structures, functions and 1345 distribution. Advances in Insect Physiology 25:267-337. 1346 Stegner ME, Richter S. 2011. Morphology of the brain in Hutchinsoniella macracantha 1347 (Cephalocarida, Crustacea). Arthropod structure & development 40:221–243. 1348 Stegner ME, Stemme T, Iliffe TM, Richter S, Wirkner CS. 2015. The brain in three crustaceans from 1349 cavernous darkness. BMC Neuroscience 16:19. doi:10.1186/s12868-015-0138-6 1350 Steinacker A. 1979. Neural and neurosecretory control of the decapod crustacean auxiliary heart. 1351 Integrative and Comparative Biology 19:67–75. doi:10.1093/icb/19.1.67 1352 Steinacker A. 1978. The anatomy of the decapod crustacean auxiliary heart. The Biological Bulletin 1353 154:497–507. doi:10.2307/1541075 1354 Stemme T, Eickhoff R, Bicker G. 2014. Olfactory projection neuron pathways in two species of marine 1355 Isopoda (Peracarida, Malacostraca, Crustacea). Tissue and Cell 46:260–263. 1356 Stemme T, Harzsch S. 2016. Remipedia. In: Schmidt-Rhaesa A, Harzsch S, Günter P (Eds.), Structure 1357 and evolution of invertebrate nervous systems. Oxford University Press, pp 522–528. 1358 Strausfeld NJ. 2012. Arthropod Brains: Evolution, Functional Elegance, and Historical Significance. 1359 Cambridge, Mass: Belknap Press. 1360 Strausfeld NJ. 2009. Brain organization and the origin of insects: an assessment. Proceedings of the 1361 Royal Society B: Biological Sciences 276:1929–1937. doi:10.1098/rspb.2008.1471 1362 Strausfeld NJ. 2005. The evolution of crustacean and insect optic lobes and the origins of chiasmata. 1363 Arthropod Structure & Development 34:235–256. 1364 Strausfeld NJ. 1998. Crustacean – Insect relationships: The use of brain characters to derive 1365 phylogeny amongst segmented invertebrates. Brain Behavior and Evolution 52:186–206. 1366 doi:10.1159/000006563 1367 Strausfeld NJ, Nassel DR. 1981. Neuroarchitectures serving compound eyes of Crustacea and insects. 1368 Handbook of sensory physiology. 1369 Sullivan JM, Beltz BS. 2005. Integration and segregation of inputs to higher-order neuropils of the 1370 crayfish brain. Journal of Comparative Neurology 481:118–126. doi:10.1002/cne.20346 1371 Sullivan JM, Beltz BS. 2004. Evolutionary changes in the olfactory projection neuron pathways of 1372 eumalacostracan crustaceans. Journal of Comparative Neurology 470:25–38. 1373 doi:10.1002/cne.11026 1374 Sullivan JM, Beltz BS. 2001. Neural pathways connecting the deutocerebrum and lateral 1375 protocerebrum in the brains of decapod crustaceans. Journal of Comparative Neurology 441:9– 1376 22. doi:10.1002/cne.1394 1377 Sullivan JM, Benton JL, Sandeman DC, Beltz BS. 2007. Adult neurogenesis: a common strategy across 1378 diverse species. Journal of Comparative Neurology 500:574–584. 1379 Sztarker J, Strausfeld N, Andrew D, Tomsic D. 2009. Neural organization of first optic neuropils in the 1380 littoral crab Hemigrapsus oregonensis and the semiterrestrial species Chasmagnathus granulatus. 1381 Journal of Comparative Neurology 513:129–150. 1382 Sztarker J, Strausfeld NJ, Tomsic D. 2005. Organization of optic lobes that support motion detection 1383 in a semiterrestrial crab. Journal of Comparative Neurology 493:396–411. 1384 Tapley DW, Buettner GR, Shick JM. 1999. Free radicals and chemiluminescence as products of the 1385 spontaneous oxidation of sulfide in seawater, and their biological implications. The Biological 1386 Bulletin 196:52–56. 1387 Thiel M, Breithaupt T. 2011. Chemical Communication in Crustaceans. Springer, New York. 1388 Thiel M, Watling L, editors. 2015. Lifestyles and Feeding Biology, Natural History of Crustacea. 1389 Oxford, New York: Oxford University Press. 1390 Tuchina O, Koczan S, Harzsch S, Rybak J, Wolff G, Strausfeld NJ, Hansson BS. 2015. Central 1391 projections of antennular chemosensory and mechanosensory afferents in the brain of the 1392 terrestrial hermit crab (Coenobita clypeatus; Coenobitidae, Anomura). Front Neuroanat 9:94. 1393 doi:10.3389/fnana.2015.00094 1394 Van Dover CL. 2000. The ecology of deep-sea hydrothermal vents. Princeton University Press. 1395 Van Dover CL, Fry B. 1994. Microorganisms as food resources at deep-sea hydrothermal vents. 1396 Limnology and Oceanography 39:51–57. 1397 Van Dover CL, Fry B, Grassle JF, Humphris S, Rona PA. 1988. Feeding biology of the shrimp Rimicaris 1398 exoculata at hydrothermal vents on the Mid-Atlantic Ridge. Marine Biology 98:209–216. 1399 doi:10.1007/BF00391196 1400 Van Dover CL, Reynolds GT, Chave AD, Tyson JA. 1996. Light at deep-sea hydrothermal vents. 1401 Geophysical research letters 23:2049–2052. 1402 Van Dover CL, Szuts EZ, Chamberlain SC, Cann JR. 1989. A novel eye in “eyeless” shrimp from 1403 hydrothermal vents of the Mid-Atlantic Ridge. Nature 337:458–460. doi:10.1038/337458a0 1404 Vitzthum H, Homberg U, Agricola H. 1996. Distribution of Dip-allatostatin I-like immunoreactivity in 1405 the brain of the locust Schistocerca gregaria with detailed analysis of immunostaining in the 1406 central complex. Journal of Comparative Neurology 369:419–437. 1407 Welsch U, Mulisch M. 2010. Romeis Mikroskopische Technik. Springer Spektrum 551:03728–14. 1408 White SN. 2000. An investigation into the characteristics and sources of light emission at deep-sea 1409 hydrothermal vents. doi:10.1575/1912/4053 1410 White SN, Chave AD, Reynolds GT, Van Dover CL. 2002. Ambient light emission from hydrothermal 1411 vents on the Mid-Atlantic Ridge. Geophysical research letters 29. 1412 Wilkens JL. 1999. Evolution of the cardiovascular system in Crustacea. Integrative and Comparative 1413 Biology 39:199–214. doi:10.1093/icb/39.2.199 1414 Williams AB, Rona PA. 1986. Two new caridean shrimps (Bresiliidae) from a hydrothermal field on 1415 the Mid-Atlantic Ridge. Journal of Crustacean Biology 6:446–462. doi:10.2307/1548184 1416 Wirkner CS, Richter S. 2010. Evolutionary morphology of the circulatory system in Peracarida 1417 (Malacostraca; Crustacea). Cladistics 26:143–167. doi:10.1111/j.1096-0031.2009.00278.x 1418 Wittfoth C, Harzsch S. 2018. Adult neurogenesis in the central olfactory pathway of 1419 dendrobranchiate and caridean shrimps: New insights into the evolution of the deutocerebral 1420 proliferative system in reptant decapods. Developmental Neurobiology 78:757–774. 1421 doi:10.1002/dneu.22596 1422 Wolff GH, Harzsch S, Hansson BS, Brown S, Strausfeld N. 2012. Neuronal organization of the 1423 hemiellipsoid body of the land hermit crab, Coenobita clypeatus: Correspondence with the 1424 mushroom body ground pattern. Journal of Comparative Neurology 520:2824–2846. 1425 doi:10.1002/cne.23059 1426 Wolff GH, Strausfeld NJ. 2015. Genealogical correspondence of mushroom bodies across 1427 invertebrate phyla. Current Biology 25:38–44. doi:10.1016/j.cub.2014.10.049 1428 Wolff GH, Thoen HH, Marshall J, Sayre ME, Strausfeld NJ. 2017. An insect-like mushroom body in a 1429 crustacean brain. eLife 6. doi:10.7554/eLife.29889 1430 Wyatt TD. 2014. Proteins and peptides as pheromone signals and chemical signatures. Animal 1431 Behaviour 97:273–280. doi:10.1016/j.anbehav.2014.07.025 1432 Yin GL, Yang JS, Cao JX, Yang WJ. 2006. Molecular cloning and characterization of FGLamide 1433 allatostatin gene from the prawn, Macrobrachium rosenbergii. Peptides 27:1241–1250. 1434 Zbinden M, Berthod C, Montagné N, Machon J, Léger N, Chertemps T, Rabet N, Shillito B, Ravaux J. 1435 2017. Comparative study of chemosensory organs of shrimp from hydrothermal vent and coastal 1436 environments. Chemical Senses 42:319–331. doi:10.1093/chemse/bjx007 1437 Zbinden M, Gallet A, Szafranski KM, Machon J, Ravaux J, Léger N, Duperron S. 2018. Blow your nose, 1438 shrimp! Unexpectedly dense bacterial communities occur on the antennae and antennules of 1439 hydrothermal vent shrimp. Frontiers in Marine Science 5. doi:10.3389/fmars.2018.00357 1440 Zbinden M, Le Bris N, Gaill F, Compère P. 2004. Distribution of bacteria and associated minerals in 1441 the gill chamber of the vent shrimp Rimicaris exoculata and related biogeochemical processes. 1442 Marine Ecology Progress Series 284:237–251. 1443 Zbinden M, Shillito B, Le Bris N, de Villardi de Montlaur C, Roussel E, Guyot F, Gaill F, Cambon- 1444 Bonavita M-A. 2008. New insigths on the metabolic diversity among the epibiotic microbial 1445 communitiy of the hydrothermal shrimp Rimicaris exoculata. Journal of Experimental Marine 1446 Biology and Ecology 359:131–140. doi:10.1016/j.jembe.2008.03.009 1447 Zeng Z, Dai L, Liu Y, Li H, Li Q, Wang J. 2002. Surface microstructures of the antennule of the white 1448 shrimp Penaeus vannamei. Journal Huazhong (Central China) Agricultural University 21:371–378. 1449 1 Aesthetascs Olfactory neuropils (ON) Neuropil Mean total Total Length glomerular Glomerular References Species (body length) volume number (µm) volume number (x10^6 (x10^3 µm3) µm3) Leptostraca

Nebalia herbstii (1.4 cm) - - 0.1 2 60 Kenning et al., 2013 Stomatopoda

Neogonodactylus oerstedii (4 cm) 80 400 - 110 70 Derby et al., 2003 Isopoda

Saduria entomon (8 cm) 40-60 240 3 34 80 Kenning and Harzsch, 2013; Pynnönen, 1985

Dendrobranchiata

Penaeus vannamei (7 cm) 280 - - - <100 Wittfoth and Harzsch, 2018; Zeng et al., 2002 Caridea

Palaemon elegans (7 cm) 280 230 120 225 530 Zbinden et al., 2017; this study* Rimicaris exoculata (6 cm) 206 170 56 155 370 Zbinden et al., 2017; this study Achelata

Panulirus argus (20-60 cm) 3000 1000 154 118 1332 Beltz et al., 2003; Grünert and Ache, 1988 Homarida

Homarus americanus (20-60 cm) 2000 600 141 592 249 Beltz et al., 2003; Guenther and Atema, 1998 Astacida

Procambarus clarkii (9 cm) 133 - 10 20 503 Beltz et al., 2003 Anomura

Birgus latro (20 cm) 1700 - 375 280 1338 Krieger et al., 2010 Coenobita clypeatus (6 cm) 519 - 120 154 799 Beltz et al., 2003 Pagurus bernhardus (3 cm) 673 - - 171 536 Tuchina et al., 2015 Brachyura

Carcinus maenas (9 cm) 200 750 - 247 - Fontaine et al., 1982; Hallberg and Skog, 2011

2 Table 1. Comparative table summarizing characteristics of aesthetascs and olfactory neuropils in several 3 malacostracan species.

4 Estimates of the animal’s body lengths are given for comparison. Carapace width is given for B. latro and C. 5 maenas, and total length is given for all other species. 6 * The palaemonid shrimp Palaemon elegans was investigated in the present study for comparison, as a species 7 closely-related to R. exoculata among the Caridea family.